Research ArticleDrought Hazard Evaluation in Boro Paddy CultivatedAreas of Western Bangladesh at Current and Future ClimateChange Conditions

Abu Reza Md Towfiqul Islam12 Shuanghe Shen1 Zhenghua Hu1 andM Atiqur Rahman13

1Collaborative Innovation Center on Forecast and Evaluation of Meteorological DisastersNanjing University of Information Science and Technology Nanjing 210044 China2Department of Disaster Management Begum Rokeya University Rangpur 5400 Bangladesh3Department of Geography and Environmental Studies University of Chittagong Chittagong 4331 Bangladesh

Correspondence should be addressed to Shuanghe Shen yqzhrnuisteducn

Received 5 September 2016 Revised 14 November 2016 Accepted 16 November 2016 Published 2 January 2017

Academic Editor Eduardo Garcıa-Ortega

Copyright copy 2017 Abu Reza Md Towfiqul Islam et al This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

Drought hazard is one of the main hindrances for sustaining food security in Bangladesh and climate change may exacerbate itin the next several decades This study aims to evaluate drought hazard at current and future climate change conditions in theBoro paddy cultivated areas of western Bangladesh using simulated climate data from the outputs of three global climate models(GCMs) based on the SRES A1B scenario for the period between 2041 and 2070 The threshold level of Standardized PrecipitationEvapotranspiration Index (SPEI) was employed to identify drought events and its probability distribution function (PDF) wasapplied to create the drought hazard index The study demonstrates that enhancement of potential evapotranspiration (PET) willsurpass that of precipitation resulting in intensified drought events in future In addition the PDFs of drought events will movethe upper tail in future period compared to the baseline The results showed that the southwestern region was more severe to thedrought hazard than the northwestern region during the period of 1984 to 2013 From the results of three GCMs in themid-centuryperiod drought hazard will slightly increase in the northwestern region and flatten with a decrease in the southwestern regionTheoutcomes will help to allocate agricultural adaptation plans under climate change condition in Bangladesh

1 Introduction

Bangladesh is one of themost natural hazard-prone countriesin the world because of its high climatic variability lowflat topography hydrogeologic setting and diverse complexgeomorphology Bangladesh has experienced the extremeclimate events in almost every year such as droughts floodstropical cyclones and storm surges causing heavy loss oflife and properties [1] Drought does immense damage tocrop production and one of the main limiting factors inthe field of agriculture Being an agriculture based countryBangladesh is struggling to be able to adapt to the changingclimate along with the challenges of a growing populationand sustaining food security The effects of climate changeon the agricultural sector are tremendous Both positive and

negative effects have occurred but the negative effects aredominated in the agricultural sector of Bangladesh [2 3]Besides climate change is likely to shift the patterns ofdrought and possibly increase the frequency and intensityof drought events in the foreseeable future [4] Westernregion of Bangladesh will be at high risk of drought hazardunder climate change conditions [5] Shahid [6] projectedan increased severity of droughts in the near future periodin Bangladesh In recent decades Bangladesh has shown anincreased drought frequency and intensity due to land usepattern changes [7] Concern among Bangladeshrsquos scientistshas increased regarding changes in precipitation potentialevapotranspiration (PET) and drought events Thus a bet-ter insight into drought frequency and intensity can helpdecision-making for agriculture allocation and adaptation to

HindawiAdvances in MeteorologyVolume 2017 Article ID 3514381 12 pageshttpsdoiorg10115520173514381

2 Advances in Meteorology

climate change especially during the dry season in the paddygrowing regions in Bangladesh

Drought is apparent over a prolong period of timegenerally a season or more in length when precipitationis below the normal levels but it is difficult to quantifydrought characteristics relative to the aspects of intensityfrequency duration and spatial extent To date more than50 drought indices have been found for recognizing droughtevents in the literature among them three are the mostwidely used drought monitoring indices such as the SPI(Standardized Precipitation Index) [8] the PDSI (PalmerDrought Severity Index) [9] and the SPEI (StandardizedPrecipitation Evapotranspiration Index) [10] The results ofthe using SPI are comparable in space and time [11] andits ability to identify various types of droughts Howeverthe SPI is based only on precipitation data which does notreflect drought conditions caused by climate variation [12]The PDSI reveals the drought effects caused by warming [13]considering the temperature data but it is unable to evaluatemultiscale drought characteristics Recently several studiesin Bangladesh have been carried out to identify droughthazard using drought indices [14ndash17] For instance Shahidand Behrawan [14] applied the SPI based on the monthlyrainfall data for the period of 1961ndash1999 from 12 weatherstations to assess the drought risk in the western part ofBangladesh Similarly Alamgir et al [16] used the SPI fromthe rainfall data during 1961ndash2010 in Bangladesh to analyzedrought events and the results revealed that the spatialcharacteristics of droughts vary widely according to seasonAbdullah [15] assessed the spatial and temporal patterns ofdrought in Bangladesh using the SPEI Rahman and Lateh[17] used SPI and monthly rainfall dataset for the periodof 1971ndash2010 to outline the drought years and severity inBangladesh However most of those studies are focused ondefining drought and their spatial temporal variations usingvarious methods with present climate condition Taking intoconsideration monthly time scale for drought monitoringand assessment under present and future climate changeconditions during the winter Boro rice growing season inwestern Bangladesh the SPEI is more suitable than otherkinds of indices for identifying drought characteristics [10]

Drought hazard has been defined as a combination ofprobability of drought frequency and intensity [18]Themorefrequent drought events with high levels of intensity canproduce the severe hazardous effects The drought hazardassessment has received recently much attention among thescientific community [19ndash21] It is necessary for disasterrisk assessment which helps in decision-making for droughtadaptation andmitigation plans [14 22] Most of the previousstudies revealed that drought frequency was considered asa basis of drought hazard but drought intensity was nottaken into account [23 24] Several studies have been doneby applying drought intensity by dividing drought intodifferent grades on the basis of values of drought index andassigned weights [17 25 26] Conversely these approacheshave limitations because artificial factors influence the gradedividing In the present study the frequency distributions ofdrought intensity within the grades are ignored The prob-ability density function (PDF) is an approach which takes

this advantage over dividing drought intensity into variousgrades Mishra et al [27] applied the PDFs of drought indexto demonstrate drought frequency and intensity successfullyA similar approach was used to assess drought hazard forChina by Q Zhang and J Zhang [21] The PDFs are muchprecise technique in comparisonwith artificial grade dividingof drought index However previous studies did not addressthe projected changes of drought events in terms of frequencyand intensity This study has taken into consideration thechanges of future drought events using the PDFs of SPEIvalues which are defined by the threshold level

A number of studies have focused on drought disaster riskon global and regional scales for present and future climatescenarios [28ndash30] It is quite important to predict futuredrought events on a local scale for drought mitigation andagriculture adaptation purposes The global climate models(GCMs) and regional climate models (RCMs) have emergedas useful tools for assessing future drought conditions underdifferent scenarios In this study three GCMs outputs wereused to evaluate future drought hazard using the IPCC SRES(Special Report on Emissions Scenario) A1B scenario for theperiod of 2041 to 2070 Burke and Brown [31] used theHadleyCenter climate model (HadCM3) as well as a multimodelensemble to evaluate the uncertainties in future drought pro-jection Chen et al [32] investigated future drought changesin China using the multimodel ensemble (MME) and aregional climatemodel (RCMs) under the SRESA1B scenarioto project a decrease in drought frequency in most partsof China On the contrary Wang et al [33] assessed futuredrought in China using a Coupled Model IntercomparisonProject Phase 5 (CMIP5) under the scenarios RCP45 andRCP85 and showed that extremedrought eventswill increasein future Recently Kim et al [20] projected future droughthazard in South Korea based on RCP85 scenario by theRCMs (HadGEM3-RA) and found that drought frequencyand intensity will become severe under future climate changeIn Bangladesh Selvaraju and Baas [34] reported that futuredrought may increase the probability of a dry year witha certain percentage of below average rainfall by 44 timeusing moderate climate change scenario and concluded thatdrought-prone areas would expand to include northwestto central regions Nevertheless a climate change projecteddrought hazard study using SPEI technique has not been welldocumented in existing literature so far especially in the caseof western Bangladesh and is the primary justification for thisstudy

ThemajorBoro paddy rice producing areas of Bangladeshare the northwestern and southwestern regions wheredrought affects 12 million hectares of cultivated Boro paddyarea during the dry season [35] The Ministry of Agricultureof Bangladesh MoA [36] reported that moderate to verysevere drought-prone areas accounted for about 569 of thecountryrsquos total net cultivated area in 2012 The drought in the1990s in northwestern Bangladesh led to a shortfall of 35mil-lion tons of rice [37] Between 1984 to 2013 drought occurredin these regions in 10 times severe droughts hit these parts in1984 1989 1991 1994 1995 1998 2000 2006 2009 and 2012[38] The average crop production reduced about 25ndash30because of the effect of drought in these regions of Bangladesh

Advances in Meteorology 3

[38] The paddy rice production inevitably incurs significantlosses from the drought hazard by threatening the countryrsquosfood security Nevertheless adaptation of drought issues inthe study area has been recently brought forward becauseof the effects of climate change increasing drought intensityand losses in the crop agricultural production It is thereforeessential to understand that drought hazard during the Boropaddy growing season and its future changes can help toformulate an effective drought preparedness and adaptationplans under climate change conditions particularly in water-scarce agricultural regions in western Bangladesh

The main objective of this study is to evaluate droughthazard at present and potential climate change conditionstaking into consideration the temporal and spatial changesof future drought events To achieve the objective first wesimulated a database of climate data (2041ndash2070) from theoutputs of three GCMs namely CGCM31 FGOALS-G10andHadGEM1models whichwere downscaled by the LARS-WG (Long Ashton Research Station-Weather Generation)model under the IPCC (Intergovernmental Panel on Cli-mate change) SRES A1B experiment Second the SPEI wasemployed to identify drought conditions by the thresholdlevel (SPEI lt minus10) Third the probability density function(PDF) of SPEI values was determined to create the droughthazard index during the current and future periods in eachsubregion of the study area The results of this study canprovide an important support for drought adaptation plansunder climate change condition

2 Data and Methods21 Study Area Description The study area is westernBangladesh which is divided into two regions northwesternregion (24∘301015840ndash26∘401015840N 88∘011015840ndash89∘901015840E) and southwest-ern region (22∘521015840ndash23∘901015840N 88∘201015840ndash90∘301015840E) (Figure 1)These two regions are considered as themain Boro paddy ricegrowing areas of Bangladesh bywhich countryrsquos food securitycan be ensured to a significant level Three kinds of ricecropping patterns such as winter (Boro) summer (Aus) andmonsoon (Aman) are usually practiced in the study area Outof these three high yield variety (HYV) Boro rice is the maincontributor of total rice production in BangladeshThis studyconsiders only the Boro paddy fields that have been severelythreatened by drought events in the last several decades TheBoro paddy is grown from January to May in these regionsthough seed bed preparation started earlier

Bangladesh enjoys the subtropical monsoon climate withlarge spatial and temporal variability and has four distinctseasons like winter spring summer and autumn About80 of the countryrsquos annual rainfall occurs during thesummer season and less than 6 rainfall occurs in thedry season Boro rice growing period The average annualrainfall of Bangladesh is about 2300mm But the annualaverage total rainfall is about 1329mm in the northwest andabout 2023mm in the southwest region [39] The averagetemperature ranges from 17 to 206∘C in the winter seasonand 269 to 311∘C in the summer According to BangladeshMeteorological Department (BMD) the summer tempera-ture rises up to 45∘C and the winter temperature falls at

Bogra

Khulna

Rangpur

Rajshahi

Dinajpur

Barisal

Faridpur

Satkhira

LegendMeteorological stationsNorthwestern regionSouthwestern region

0 40 80 120 16020(km)

N

Bay of Bengal

26∘N

25∘N

24∘N

23∘N

22∘N

21∘N

20∘N

92∘E91

∘E90∘E89

∘E88∘E

Figure 1 The location map showing meteorological stations andthe extent of the northwestern and southwestern regions in westernBangladesh

5∘C in the northwest region These regions experience twoextremities that clearly differentiate the climatic behaviorsfrom the rest of the country [40]

The southwestern part has experienced more frequentextreme events than the northwestern part and also the restof the country The northwestern region has been faced withrecurrent mean rainfall 6494mm in the last 30-year period(1984ndash2013) while the southwestern part has experiencedmean rainfall 7194mm during winter rice growing season(Table 1) It can be seen from Table 1 that northwesternpart prevails comparatively higher rainfall variability thansouthwestern part which is highly significant The varia-tions in temperature are not so significant in these partsThe northwestern subregions prevail generally warmer andhumid climatic conditions than the southwestern subregionCropping intensity has increased almost double since lastthree decades in the northwestern region than southwesternregion Northwestern region belongs to red soil characteris-tics having low water holding capacity which differentiatesit from the southwestern region Western Bangladesh alsovaries from altitude for example altitude is the lowest(4m) in Barishal subregion of the southwestern region andaltitude is the highest (37m) in Dinajpur subregion of thenorthwestern region compared to the rest of the country

4 Advances in Meteorology

Table 1 The variations in climatic parameters during the winterBoro paddy growing season (1984ndash2013) in western Bangladesh

Region Subregion 119879max 119879min Precipitation

Northwest

Dinajpur 2926 1712 6221Rangpur 2866 1726 8897Bogra 3009 1829 6258

Rajshahi 3131 1782 4616Mean 2983 176225 6498

Southwest

Faridpur 3071 1894 8021Khulna 3142 1961 6623Satkhira 3161 1974 6349Barisal 3075 1932 7771Mean 311225 194025 7191

However the severity of drought in southwestern part ismoderate compared to northwestern region during the drywinter season [38] So drought hazard evaluation underclimate change condition has been carried out in varyingclimatic characteristics western Bangladesh varies latitudeand altitude but longitude is almost similar

22 Datasets Generation and Statistical Downscaling In thisstudy we divided the time domain into two periods cur-rent (1984ndash2013) and near future (2041ndash2070) for potentialdrought hazard analysis According to IPCC [41] the timeperiod of 1984 to 2013 has considered the warmest 30-yearperiod for the last 100 years For this reason the dailyclimate data of 30-year period (1984ndash2013) was used asan observed period The historical climate data (eg min-imum and maximum temperature and precipitation) wereobtained from the Bangladesh Meteorological Department(BMD) There are 6 weather stations in the northwesternregion and 7 in the southwestern region Although BMDhas 13 meteorological stations in western Bangladesh fewstations for example Sayedpur (northwestern region) andMongla and Chuadanga stations (southwestern region) donot have long-term observed data archives because theseare newly established after 1991 and have missing data forsignificant periods Thus data of total 8 meteorologicalstations in the two regions for a 30-year period (1984ndash2013)were taken into consideration The time period of 2041ndash2070 at centered 2055s is regarded as a planning horizonof crop agricultural productivity The present study chose toevaluate mid-century future climate (2041ndash2070) due to thelimited ability for developing meaningful crop agriculturalmanagement strategies relevant to end of the century climateprojections The IPCC Data Distribution Center providesthe three GCMs results based on the SRES A1B scenarios(httpwwwipcc-dataorg) The GCMs that provide a resultof the A1B scenario applied for the IPCC AR4 were selectedCGCM31 (Canada) FGOALS-G10 (China) and HadGEM1(UK) The outputs from three GCMs climate models areemployed in this study as summarized in Table 2 TheA1B scenario (balanced across energy sources) represents amedium greenhouse gas emission

The LARS-WG is a stochastic weather generator devel-oped by Semenov and Barrow [42] which generates precipi-tation andmaximum andminimum temperature for the timeperiod of 1984 to 2013 and future period of 2041 to 2070 It isa statistical downscaling method which is used to calibratethe climate model outputs with a statistical relationshipbetween the GCMs result and the observed result Semenovet al [43] applied WGEN (weather generator) and LARS-WG method in different climatic zones in Europe and Asiaand found that the LARS-WG was more accurate than theWGEN method The parameters of LARS-WG model weregenerated from 8 representative weather stations based onthe major cultivated Boro rice areas in western BangladeshThe LARS-WGmodel was downloaded from the departmentof computational and systems biology Rothamsted researchUK website (httpwwwrothamstedacuk) More detaileddescription of the LARS-WG modeling procedure can bereferred to Semenov and Barrow [42] It is mentionablethat the observed current climate (1984ndash2013) is significantlydifferent from the climate in the past years for example1961ndash1990 However the inverse distance weighting (IDW)interpolation technique was applied to assess drought hazardunder climate change in South Korea [19] and in Bangladesh[17] and was also adopted for use in the study The IDWinterpolation technique is simple and in-built within ArcGIS(version 102) Additionally the main advantage of thisinterpolation technique is fast to compute the interpolatedvalues

23 SPEI and Threshold Level The Standardized Precipita-tion Evapotranspiration Index (SPEI) was used to identifythe drought characteristics in this study The SPEIrsquos mainadvantage lies in its ability to detect the onset and spatial andtemporal changes of drought consistently it is suggested foroperational drought monitoring studies worldwide [44 45]However the Standardized Precipitation Index (SPI) does notconsider temperature data it has a limitation which is unableto reflect the changes in water demand such as the changes inwater budget like precipitation and PET by the warmingTheSPEI rectified this criticism of the SPI considering potentialevapotranspiration (PET)

In the present study the following four steps and specificsettings are used to calculate the SPEI [46] Firstly themonthly potential evapotranspiration (PET) is computedbased on the data of monthly minimum and maximum tem-perature by using the Thornthwaite equation [47] Secondlya simple monthly water balance is expressed as the differencebetween monthly precipitation (119875) and PETThirdly a three-parameter log-logistic distribution is used to fit the data seriesof monthly water balance Finally the original values arenormalized to obtain standardized units that are comparablein space and time as the SPEI The updated version of the RSPEI package (httpcranr-projectorgwebpackagesSPEI)was used to estimate the SPEI in the study The SPEI wascomputed at monthly time scale and the smaller negativeSPEI means the most serious drought events will be occur-ring The negative SPEIs are related to the dry conditiona drought event is defined when the SPEI is continuouslynegative and reaches a value of ldquominus10rdquo or less [10] So it is

Advances in Meteorology 3

[38] The paddy rice production inevitably incurs significantlosses from the drought hazard by threatening the countryrsquosfood security Nevertheless adaptation of drought issues inthe study area has been recently brought forward becauseof the effects of climate change increasing drought intensityand losses in the crop agricultural production It is thereforeessential to understand that drought hazard during the Boropaddy growing season and its future changes can help toformulate an effective drought preparedness and adaptationplans under climate change conditions particularly in water-scarce agricultural regions in western Bangladesh

The main objective of this study is to evaluate droughthazard at present and potential climate change conditionstaking into consideration the temporal and spatial changesof future drought events To achieve the objective first wesimulated a database of climate data (2041ndash2070) from theoutputs of three GCMs namely CGCM31 FGOALS-G10andHadGEM1models whichwere downscaled by the LARS-WG (Long Ashton Research Station-Weather Generation)model under the IPCC (Intergovernmental Panel on Cli-mate change) SRES A1B experiment Second the SPEI wasemployed to identify drought conditions by the thresholdlevel (SPEI lt minus10) Third the probability density function(PDF) of SPEI values was determined to create the droughthazard index during the current and future periods in eachsubregion of the study area The results of this study canprovide an important support for drought adaptation plansunder climate change condition

2 Data and Methods21 Study Area Description The study area is westernBangladesh which is divided into two regions northwesternregion (24∘301015840ndash26∘401015840N 88∘011015840ndash89∘901015840E) and southwest-ern region (22∘521015840ndash23∘901015840N 88∘201015840ndash90∘301015840E) (Figure 1)These two regions are considered as themain Boro paddy ricegrowing areas of Bangladesh bywhich countryrsquos food securitycan be ensured to a significant level Three kinds of ricecropping patterns such as winter (Boro) summer (Aus) andmonsoon (Aman) are usually practiced in the study area Outof these three high yield variety (HYV) Boro rice is the maincontributor of total rice production in BangladeshThis studyconsiders only the Boro paddy fields that have been severelythreatened by drought events in the last several decades TheBoro paddy is grown from January to May in these regionsthough seed bed preparation started earlier

Bangladesh enjoys the subtropical monsoon climate withlarge spatial and temporal variability and has four distinctseasons like winter spring summer and autumn About80 of the countryrsquos annual rainfall occurs during thesummer season and less than 6 rainfall occurs in thedry season Boro rice growing period The average annualrainfall of Bangladesh is about 2300mm But the annualaverage total rainfall is about 1329mm in the northwest andabout 2023mm in the southwest region [39] The averagetemperature ranges from 17 to 206∘C in the winter seasonand 269 to 311∘C in the summer According to BangladeshMeteorological Department (BMD) the summer tempera-ture rises up to 45∘C and the winter temperature falls at

Bogra

Khulna

Rangpur

Rajshahi

Dinajpur

Barisal

Faridpur

Satkhira

LegendMeteorological stationsNorthwestern regionSouthwestern region

0 40 80 120 16020(km)

N

Bay of Bengal

26∘N

25∘N

24∘N

23∘N

22∘N

21∘N

20∘N

92∘E91

∘E90∘E89

∘E88∘E

Figure 1 The location map showing meteorological stations andthe extent of the northwestern and southwestern regions in westernBangladesh

5∘C in the northwest region These regions experience twoextremities that clearly differentiate the climatic behaviorsfrom the rest of the country [40]

The southwestern part has experienced more frequentextreme events than the northwestern part and also the restof the country The northwestern region has been faced withrecurrent mean rainfall 6494mm in the last 30-year period(1984ndash2013) while the southwestern part has experiencedmean rainfall 7194mm during winter rice growing season(Table 1) It can be seen from Table 1 that northwesternpart prevails comparatively higher rainfall variability thansouthwestern part which is highly significant The varia-tions in temperature are not so significant in these partsThe northwestern subregions prevail generally warmer andhumid climatic conditions than the southwestern subregionCropping intensity has increased almost double since lastthree decades in the northwestern region than southwesternregion Northwestern region belongs to red soil characteris-tics having low water holding capacity which differentiatesit from the southwestern region Western Bangladesh alsovaries from altitude for example altitude is the lowest(4m) in Barishal subregion of the southwestern region andaltitude is the highest (37m) in Dinajpur subregion of thenorthwestern region compared to the rest of the country

4 Advances in Meteorology

Table 1 The variations in climatic parameters during the winterBoro paddy growing season (1984ndash2013) in western Bangladesh

Region Subregion 119879max 119879min Precipitation

Northwest

Dinajpur 2926 1712 6221Rangpur 2866 1726 8897Bogra 3009 1829 6258

Rajshahi 3131 1782 4616Mean 2983 176225 6498

Southwest

Faridpur 3071 1894 8021Khulna 3142 1961 6623Satkhira 3161 1974 6349Barisal 3075 1932 7771Mean 311225 194025 7191

However the severity of drought in southwestern part ismoderate compared to northwestern region during the drywinter season [38] So drought hazard evaluation underclimate change condition has been carried out in varyingclimatic characteristics western Bangladesh varies latitudeand altitude but longitude is almost similar

22 Datasets Generation and Statistical Downscaling In thisstudy we divided the time domain into two periods cur-rent (1984ndash2013) and near future (2041ndash2070) for potentialdrought hazard analysis According to IPCC [41] the timeperiod of 1984 to 2013 has considered the warmest 30-yearperiod for the last 100 years For this reason the dailyclimate data of 30-year period (1984ndash2013) was used asan observed period The historical climate data (eg min-imum and maximum temperature and precipitation) wereobtained from the Bangladesh Meteorological Department(BMD) There are 6 weather stations in the northwesternregion and 7 in the southwestern region Although BMDhas 13 meteorological stations in western Bangladesh fewstations for example Sayedpur (northwestern region) andMongla and Chuadanga stations (southwestern region) donot have long-term observed data archives because theseare newly established after 1991 and have missing data forsignificant periods Thus data of total 8 meteorologicalstations in the two regions for a 30-year period (1984ndash2013)were taken into consideration The time period of 2041ndash2070 at centered 2055s is regarded as a planning horizonof crop agricultural productivity The present study chose toevaluate mid-century future climate (2041ndash2070) due to thelimited ability for developing meaningful crop agriculturalmanagement strategies relevant to end of the century climateprojections The IPCC Data Distribution Center providesthe three GCMs results based on the SRES A1B scenarios(httpwwwipcc-dataorg) The GCMs that provide a resultof the A1B scenario applied for the IPCC AR4 were selectedCGCM31 (Canada) FGOALS-G10 (China) and HadGEM1(UK) The outputs from three GCMs climate models areemployed in this study as summarized in Table 2 TheA1B scenario (balanced across energy sources) represents amedium greenhouse gas emission

The LARS-WG is a stochastic weather generator devel-oped by Semenov and Barrow [42] which generates precipi-tation andmaximum andminimum temperature for the timeperiod of 1984 to 2013 and future period of 2041 to 2070 It isa statistical downscaling method which is used to calibratethe climate model outputs with a statistical relationshipbetween the GCMs result and the observed result Semenovet al [43] applied WGEN (weather generator) and LARS-WG method in different climatic zones in Europe and Asiaand found that the LARS-WG was more accurate than theWGEN method The parameters of LARS-WG model weregenerated from 8 representative weather stations based onthe major cultivated Boro rice areas in western BangladeshThe LARS-WGmodel was downloaded from the departmentof computational and systems biology Rothamsted researchUK website (httpwwwrothamstedacuk) More detaileddescription of the LARS-WG modeling procedure can bereferred to Semenov and Barrow [42] It is mentionablethat the observed current climate (1984ndash2013) is significantlydifferent from the climate in the past years for example1961ndash1990 However the inverse distance weighting (IDW)interpolation technique was applied to assess drought hazardunder climate change in South Korea [19] and in Bangladesh[17] and was also adopted for use in the study The IDWinterpolation technique is simple and in-built within ArcGIS(version 102) Additionally the main advantage of thisinterpolation technique is fast to compute the interpolatedvalues

23 SPEI and Threshold Level The Standardized Precipita-tion Evapotranspiration Index (SPEI) was used to identifythe drought characteristics in this study The SPEIrsquos mainadvantage lies in its ability to detect the onset and spatial andtemporal changes of drought consistently it is suggested foroperational drought monitoring studies worldwide [44 45]However the Standardized Precipitation Index (SPI) does notconsider temperature data it has a limitation which is unableto reflect the changes in water demand such as the changes inwater budget like precipitation and PET by the warmingTheSPEI rectified this criticism of the SPI considering potentialevapotranspiration (PET)

In the present study the following four steps and specificsettings are used to calculate the SPEI [46] Firstly themonthly potential evapotranspiration (PET) is computedbased on the data of monthly minimum and maximum tem-perature by using the Thornthwaite equation [47] Secondlya simple monthly water balance is expressed as the differencebetween monthly precipitation (119875) and PETThirdly a three-parameter log-logistic distribution is used to fit the data seriesof monthly water balance Finally the original values arenormalized to obtain standardized units that are comparablein space and time as the SPEI The updated version of the RSPEI package (httpcranr-projectorgwebpackagesSPEI)was used to estimate the SPEI in the study The SPEI wascomputed at monthly time scale and the smaller negativeSPEI means the most serious drought events will be occur-ring The negative SPEIs are related to the dry conditiona drought event is defined when the SPEI is continuouslynegative and reaches a value of ldquominus10rdquo or less [10] So it is

4 Advances in Meteorology

Table 1 The variations in climatic parameters during the winterBoro paddy growing season (1984ndash2013) in western Bangladesh

Region Subregion 119879max 119879min Precipitation

Northwest

Dinajpur 2926 1712 6221Rangpur 2866 1726 8897Bogra 3009 1829 6258

Rajshahi 3131 1782 4616Mean 2983 176225 6498

Southwest

Faridpur 3071 1894 8021Khulna 3142 1961 6623Satkhira 3161 1974 6349Barisal 3075 1932 7771Mean 311225 194025 7191

However the severity of drought in southwestern part ismoderate compared to northwestern region during the drywinter season [38] So drought hazard evaluation underclimate change condition has been carried out in varyingclimatic characteristics western Bangladesh varies latitudeand altitude but longitude is almost similar

22 Datasets Generation and Statistical Downscaling In thisstudy we divided the time domain into two periods cur-rent (1984ndash2013) and near future (2041ndash2070) for potentialdrought hazard analysis According to IPCC [41] the timeperiod of 1984 to 2013 has considered the warmest 30-yearperiod for the last 100 years For this reason the dailyclimate data of 30-year period (1984ndash2013) was used asan observed period The historical climate data (eg min-imum and maximum temperature and precipitation) wereobtained from the Bangladesh Meteorological Department(BMD) There are 6 weather stations in the northwesternregion and 7 in the southwestern region Although BMDhas 13 meteorological stations in western Bangladesh fewstations for example Sayedpur (northwestern region) andMongla and Chuadanga stations (southwestern region) donot have long-term observed data archives because theseare newly established after 1991 and have missing data forsignificant periods Thus data of total 8 meteorologicalstations in the two regions for a 30-year period (1984ndash2013)were taken into consideration The time period of 2041ndash2070 at centered 2055s is regarded as a planning horizonof crop agricultural productivity The present study chose toevaluate mid-century future climate (2041ndash2070) due to thelimited ability for developing meaningful crop agriculturalmanagement strategies relevant to end of the century climateprojections The IPCC Data Distribution Center providesthe three GCMs results based on the SRES A1B scenarios(httpwwwipcc-dataorg) The GCMs that provide a resultof the A1B scenario applied for the IPCC AR4 were selectedCGCM31 (Canada) FGOALS-G10 (China) and HadGEM1(UK) The outputs from three GCMs climate models areemployed in this study as summarized in Table 2 TheA1B scenario (balanced across energy sources) represents amedium greenhouse gas emission

The LARS-WG is a stochastic weather generator devel-oped by Semenov and Barrow [42] which generates precipi-tation andmaximum andminimum temperature for the timeperiod of 1984 to 2013 and future period of 2041 to 2070 It isa statistical downscaling method which is used to calibratethe climate model outputs with a statistical relationshipbetween the GCMs result and the observed result Semenovet al [43] applied WGEN (weather generator) and LARS-WG method in different climatic zones in Europe and Asiaand found that the LARS-WG was more accurate than theWGEN method The parameters of LARS-WG model weregenerated from 8 representative weather stations based onthe major cultivated Boro rice areas in western BangladeshThe LARS-WGmodel was downloaded from the departmentof computational and systems biology Rothamsted researchUK website (httpwwwrothamstedacuk) More detaileddescription of the LARS-WG modeling procedure can bereferred to Semenov and Barrow [42] It is mentionablethat the observed current climate (1984ndash2013) is significantlydifferent from the climate in the past years for example1961ndash1990 However the inverse distance weighting (IDW)interpolation technique was applied to assess drought hazardunder climate change in South Korea [19] and in Bangladesh[17] and was also adopted for use in the study The IDWinterpolation technique is simple and in-built within ArcGIS(version 102) Additionally the main advantage of thisinterpolation technique is fast to compute the interpolatedvalues

23 SPEI and Threshold Level The Standardized Precipita-tion Evapotranspiration Index (SPEI) was used to identifythe drought characteristics in this study The SPEIrsquos mainadvantage lies in its ability to detect the onset and spatial andtemporal changes of drought consistently it is suggested foroperational drought monitoring studies worldwide [44 45]However the Standardized Precipitation Index (SPI) does notconsider temperature data it has a limitation which is unableto reflect the changes in water demand such as the changes inwater budget like precipitation and PET by the warmingTheSPEI rectified this criticism of the SPI considering potentialevapotranspiration (PET)

In the present study the following four steps and specificsettings are used to calculate the SPEI [46] Firstly themonthly potential evapotranspiration (PET) is computedbased on the data of monthly minimum and maximum tem-perature by using the Thornthwaite equation [47] Secondlya simple monthly water balance is expressed as the differencebetween monthly precipitation (119875) and PETThirdly a three-parameter log-logistic distribution is used to fit the data seriesof monthly water balance Finally the original values arenormalized to obtain standardized units that are comparablein space and time as the SPEI The updated version of the RSPEI package (httpcranr-projectorgwebpackagesSPEI)was used to estimate the SPEI in the study The SPEI wascomputed at monthly time scale and the smaller negativeSPEI means the most serious drought events will be occur-ring The negative SPEIs are related to the dry conditiona drought event is defined when the SPEI is continuouslynegative and reaches a value of ldquominus10rdquo or less [10] So it is

Advances in Meteorology 5

Table 2 Summary of observed and simulated climate data and three GCMs outputs used in this study

Category Period Source Climate model Spatial resolution(lat times long)

Observed data 1984ndash2013 BMD (Bangladesh Meteorological Department) Historical data mdash

Simulated data 2041ndash2070 IPCC Data Distribution Centre (A1B scenario)CGCM31 (Canada) 28∘ times 28∘

FGOALS-G10 (China) 28∘ times 28∘HadGEM1 (UK) 13∘ times 19∘

JanuaryFebruary

00

01

02

03

04

PDF

10 2 3minus2 minus1minus3

SPEI

Figure 2 The PDFs of SPEI values in January and Februaryduring the observed period of 1984ndash2013 in Rangpur northwesternBangladesh

assumed that less than ldquominus10rdquo is considered as the thresholdlevel and drought events start at this level lower than minus10 inmonthly SPEI values (Figure 2) Two reasons for the choosingof this threshold level are as follos (1) the SPEI (ltminus10) is agood indication that significant impacts can occur in cropagriculture and possibly other sectors and (2) the thresholdlevel of SPEI (ltminus1) indicates the start of agricultural droughtevents in the crop growing season and is also used to identifyand explore drought periods Similarly Kim et al [20] usedminus10 as a threshold level when they evaluated future droughthazard of South Korea using SPEI technique

24 Drought Hazard Index Drought hazard is the productof frequency and intensity of drought events The SPEI isemployed to quantify drought hazard during the current andnear future periods Next the probability density function(PDF) of SPEI values was determined during these periods ineach subregion of the study area We computed the droughthazard (DH) by using the following equation

DH = intminus35

minus10

119868 (SPEI) times 119865 (SPEI) (1)

where 119868(SPEI) values denote the intensity of drought and119865(SPEI) is the frequency of the SPEI The justification forthis equationrsquos over widely used other drought hazard indexlies its ability to take advantage the PDFs of SPEI values interms of frequency and intensity of drought period to buildan index which avoids dividing the drought into various

grades Another advantage is that this equation measures theintensity and frequency of drought events precisely at anylocation and any time scale

The PDFs of SPEI values in January and February duringthe period of 1984ndash2013 in Rangpur subregion northwesternBangladesh are shown in Figure 2 as an example It demon-strates that drought frequencies (SPEI lt minus10) are almostsame during these two months but drought intensities varywhen the SPEI values distributed fromminus10 tominus20 in Januaryand the peak isminus10 in February Hence drought hazard (DH)value in January is higher compared to in February becausedrought intensities are greater in January than in FebruaryHowever artificial factors often influence drought gradedividing in drought hazard studies For instance variousweights are assigned to several intensity grades in droughtintensity analysis which sometime mislead the accuracy ofthe results For that reason the frequency distributions ofdrought intensity within the grades are ignored in the presentstudy

This study covers only the Boro paddy growing seasonfrom January to May where the DH is calculated as theaverage monthly DHs values We constructed a database ofobserved period (1984ndash2013) and future period (2041ndash2070)from the three GCMs results which were downscaled by theLARS-WG model under the SRES A1B scenario a 120-yeardata sequence of each subregion was used to quantify theSPEI values

25 Validation of Climate Datasets and the LARS-WG ModelPerformance The downscaled LARS-WGmodel was appliedto generate simulated climate data for the time period of2041 to 2070 using the three GCMs results These climatesimulations were verified with observed data (1984ndash2013) Toevaluate the statistically significant differences between theobserved and simulated climate data Q-test was performedincluding t-test and F-test using the LARS-WG downscalingtechnique Table 3 shows the results ofQ-test for precipitationand minimum and maximum temperature at 8 meteoro-logical stations in the study area The estimated parametersof three climate variables were assessed by Q-test at 5significance level for almost all stations During the Bororice growing period from January to May the three climatevariables at 7-8 stations were verified by Q-test with a 95confidence level

Figure 3 uses Rangpur in the northwestern region andSatkhira in the southwestern region as examples to measurethe downscaling model performance Blue lines denote theobservation data and red lines indicate simulation dataThe LARS-WG model displayed high accuracy projection of

6 Advances in Meteorology

Table 3The results ofQ-test for precipitation and minimum and maximum temperature at 8 meteorological stations in the study area (unitnumber of station)

Variables P value Jan Feb Mar Apl May Jun Jul Aug Sep Oct Nov Dec

Precipitation lt005 1 0 0 0 0 2 1 0 1 0 3 1ge005 7 8 8 8 8 6 7 8 7 8 5 7

Minimum temperature lt005 0 1 0 0 0 0 0 0 0 0 0 0ge005 8 7 8 8 8 8 8 8 8 8 8 8

Maximum temperature lt005 0 0 0 0 0 0 0 0 0 0 0 0ge005 8 8 8 8 8 8 8 8 8 8 8 8

0 200 600Precipitation (mm)

ObservedSimulated

00000

00010

00020

Den

sity

minus200

(a)

ObservedSimulated

000

005

010

015

Den

sity

25 30 3520Max temp (∘C)

(b)

ObservedSimulated

000

002

004

006

Den

sity

10 20 305Min temp (∘C)

(c)

ObservedSimulated

00000

00010

00020

00030

Den

sity

0 200 400minus200

Precipitation (mm)

(d)

ObservedSimulated

000

005

010

015

Den

sity

30 35 4025Max temp (∘C)

(e)

ObservedSimulated

10 15 20 25 30 355000

002

004

006

008D

ensit

y

Min temp (∘C)

(f)

Figure 3 Comparison of observed and simulated climate variables in the paddy growing season (andashc) show precipitation and maximumand minimum temperature in Rangpur subregion northwestern Bangladesh (dndashf) represent precipitation and maximum and minimumtemperature in Satkhira subregion southwestern Bangladesh

monthly maximum and minimum temperature in case sta-tions (Figure 3) The justification is that these two represen-tative stations are used as an example because these stationsexhibit rainfall variability that are statistically significantTherainfall patterns of two stations are changing increasinglyand annual rainfall shows a declining trend [14] The modelshowed the better performance in the southwestern than thenorthwestern region for precipitation The linear regressionanalyses confirmed a very high correlation between theobserved and simulated precipitation (1199032 = 097 and 098)at 005 significance level in both regions Furthermorethe correlation coefficients (1199032) between the simulated and

observed temperature are around 099 The comparisonbetween simulated and observed data cannot guarantee theirvalidity of future climate because global warming plays a vitalrole in the near future climate change conditions which maynot be identical to the current climate

3 Result and Discussions

The ultimate goal of this study is to evaluate drought hazardin typical Boro paddy cultivated areas of western Bangladeshat present and future climate change conditions First wewill show the temporal variations in precipitation potential

Advances in Meteorology 7

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

RajshahiDinajpur Rangpur BograSubregions

50

100

150

200Pr

ecip

itatio

n (m

m)

(a) Northwestern region

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

Faridpur BarisalSatkhiraKhulnaSubregions

50

100

150

200

Prec

ipita

tion

(mm

)

(b) Southwestern region

Figure 4 The temporal variations in average precipitation during the paddy growing season under observed and future climate changeconditions in western Bangladesh (a) The changes of precipitation in the northwestern region (b) the changes of precipitation in thesouthwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 models respectively

evapotranspiration (PET) and drought events (SPEI lt minus1)under climate change during the Boro rice growing seasonin the designated areas This study illustrates the processesthat exaggerate drought hazard in the future Hereafterwe outline changes in precipitation PET and SPEI lt minus1from the observed (1984ndash2013) to future periods (2041ndash2070)in western Bangladesh The A1B scenario is used underthree GCMs outputs which are denoted as future model1 (CGCM31) future model 2 (FGOALS-G10) and futuremodel 3 (HadGEM1)

31 Changes of Precipitation PET and Drought Events underClimate Change

311 Changes in the Precipitation Figure 4 presents the tem-poral variation in average precipitation during the growingseason (January toMay) for the current and future periods inthemajor Boro paddy cultivated areas in western BangladeshAll the subregions showed that the average precipitation dur-ing the Boro rice growing season in northwestern Bangladeshwere 9495mm (future model 1) 7877mm (future model2) and 8163mm (future model 3) by the period of 2041ndash2070 compared to 6720mm in the current period of 1984ndash2013The highest average seasonal precipitationwas observedin Rangpur (14091mm) using future model 1 while thelowest seasonal precipitation was found in Bogra (5688mm)under future model 2 However Dinajpur demonstrated adecrease in precipitation by the period of 2041ndash2070 underfuture model 2 and future model 3 compared to the currentperiod Conversely in the case of the southwestern regionall subregions exhibited that average seasonal precipitationwas 9447mm 7653mm and 7226mm respectively in nearfuture period and 7218mm in the baseline periodThe largestaverage seasonal precipitation was in Faridpur (11218mm)under future model 1 while the smallest average seasonalprecipitation was in Satkhira (5859mm) using future model3 (Figure 4) From the results it is clear that the temporalvariations in precipitation are higher in the northwesternregion than the southwestern Bangladesh The increasingtendency of precipitation under potential climate changesmay be due to high variations in rainy days amounts ofprecipitation and intensity in the Boro rice growing season in

Bangladesh In addition the huge water vapor transport fromthe Bay of Bengal and enhanced moisture convergence maybe another cause of increasing precipitation in future [33]

312 Changes in the Potential Evapotranspiration Figure 5exhibits future change in average PET in the Boro ricegrowing season for major subregions of western Bangladeshat the current and future periods under A1B scenario Theaverage seasonal PET during January to May in the north-western region was 11141mm in the observed period andincrease in the near future about 14763mm (future model 1)13487mm (future model 2) and 14542mm (future model3) respectively by the period of 2041 to 2070 The highestaverage seasonal PET was detected in Rajshahi (17035mm)under future model 3 whereas the lowest average PET wasobserved in Rangpur (12041mm) using future model 2(Figure 5)

On the other hand average PET of four subregions inthe southwestern Bangladesh was 13327mm for the currentperiod and increase in 17389 15922mm and 18304mmrespectively for future period using the three GCMs resultsThe largest seasonal average PET was noticed in Satkhira(19511mm) under future model 3 while the smallest PETwas observed in Faridpur (15043mm) using future model 2(Figure 5)The results show that the PET variations in north-western region are much larger compared to that in south-western region Furthermore the discrepancies between thePET changes using the three GCMs become more noticeablein the mid-century period This finding is consistent withthe previous results of Shahid [6] that showed an increaserate of potential evapotranspiration (PET) in northwesternBangladesh The widespread increase in PET is expected torise in temperature and amounts of solar radiation underclimate change conditions that dominate in the western partof Bangladesh Although the tendency of precipitation andPET are both upward trends the increasing rate of PETexceeds that of precipitation indicating that the increasingprecipitation is not offsetting effects of higher evaporationThe rise in global temperature and surface warming playsa dominant role in the future PET enhancement The studyimplies that enhancement of PET in the study area willaggravate drought frequency and intensity in the near future

8 Advances in Meteorology

Dinajpur RangpurSubregions

Bogra Rajshahi

100120140160180

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(a) Northwestern region

Faridpur KhulnaSubregions

BarisalSatkhira

100120140160180200

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(b) Southwestern region

Figure 5The temporal variations in average PET during the growing season under observed and future climate change conditions in westernBangladesh (a) The changes of potential evapotranspiration (PET) in northwestern region (b) the changes of potential evapotranspiration(PET) in southwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 modelsrespectively

313 Changes in the Drought Events Figure 6 shows theprobability density function (PDF) of the SPEI by thethreshold level during the Boro paddy cultivating season innorthwestern and southwestern regions under current andfuture climate change conditions at monthly time scale Thetotal SPEI valuesltminus1 during January toMay at current period(1984ndash2013) are mainly normally distributed according to themathematical principle of the SPEI procedure The resultsreveal that the PDFs become broader with future time andpeak probability distribution will increase in the near futurefor the period between 2041 and 2070 in comparison to theobserved period For instance Dinajpur subregion shows asevere increased drought events in terms of intensity and fre-quency under A1B scenario using future model 3 (HadGEM1GCMs) In addition Rajsahhi and Bogra subregions exhibitmore drought intensity in potential future climate changecompared to the current period The results show that thePDFs will move the upper tail continuously toward the mid-century period and also shift to some extent in left side(Figure 6)

Figure 6 exhibits the trend of a higher increase in the SPEIltminus1 in Khulna and Satkhira in future period compared to theothers subregions in the southwestern Bangladesh Compar-ing to drought events in the southwestern region with thenorthwestern region it is obvious that the PDFs of monthlytotal SPEI values below minus1 in the Boro paddy cultivatedarea will change drastically in the near future The futuremodel 1 (CGCM31 GCMs) and future model 2 (FGOALS-G10 GCMs) indicate that frequency of extreme droughtevents (SPEI lt minus1) is expected to increase dramatically in thewestern Bangladesh It is worth mentioning that the risk ofextreme drought events tends to increase in future climateconditions Besides drought frequency and intensity implya steady increase from the current to mid-century periodsbecause the PDFs of SPEI values shift more than one and ahalf standard deviation peaking at minus15 in all the subregionsof western Bangladesh during that periodThe results of threeGCMs show that the upper tail of PDFswill become thicker infuture period (Figure 6)The study indicates that the PDFs ofdrought events will continue tomove upward in future periodcompared to the current period The results demonstrate

that the northwestern region will increase a severe droughtfrequency and levels of intensity in comparison with thesouthwestern region in the near future

32 Spatial Variation in Drought Hazard under ClimateChange The spatial variation in drought hazard (DH) foreach subregion in western Bangladesh is computed basedon the LARS-WG downscaled precipitation and temperaturedata for the period between 2041 and 2070 Figure 7 displaysthe spatial distribution in DH during the Boro rice growingseason using the results of future 1 (CGCM31 GCMs) future2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs)models respectively

The DH value in the northwestern region ranges from180 to 261 for the current period with the highest DHvalue distributed in Rajshahi subregion and the lowest DHvalue exhibited in Rangpur subregion Conversely the DHvalue in the southwestern region ranges between 239 and376 during the baseline period (1984ndash2013) The highest DHvalue was found in Satkhira subregion while the lowest DHvalue was distributed in Faridpur subregion The obtainedresults indicated that the amplitude of DH was much largerin the southwestern region than the northwestern regionduring the period of 1984 to 2013 The findings are consistentwith the earlier studies of Abedin et al [48] and Abdullah[15] they reported that southwestern coastal region is muchdrought-prone in comparison to the rest of the country Butthe findings of Shahid and Behrawan [14] Alamgir et al [16]and Islam et al [49] disagree with the present observationof drought hazard analysis They found that drought poseda high risk to northern and northwestern part than southernand southwestern parts of Bangladesh in the current periodShahid and Behrawan [14] used onlymonthly rainfall data forthe period of 1961ndash1999 with old datasets in the western partof Bangladesh and did not consider temperature data Thepresent study has taken into consideration both precipitationand temperature data (1984ndash2070) under current and futureclimate change conditions in the winter Boro paddy growingseason of western Bangladesh This is the reason that thefindings are not consistent with those previous studies In thenear future period (2041ndash2070) especially the northwestern

Advances in Meteorology 9

ObservedFuture 1

Future 2Future 3

0 2minus2minus4

SPEI

00

02

04

06

Den

sity

(a) Dinajpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

12

Den

sity

(b) Rangpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

Den

sity

(c) Bogra

ObservedFuture 1

Future 2Future 3

0 1minus4 minus2minus6

SPEI

00

04

Den

sity

(d) Rajshahi

ObservedFuture 1

Future 2Future 3

minus4 0minus6 minus2

SPEI

00

02

04

06

Den

sity

(e) Faridpur

ObservedFuture 1

Future 2Future 3

minus4 minus2 0 2minus6

SPEI

00

02

04

06

Den

sity

(f) Khulna

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

03

06

Den

sity

(g) Satkhira

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

02

04

Den

sity

(h) Barisal

Figure 6 The probability density function (PDF) of SPEI values by the threshold level during January to May for the A1B scenario usingthe three GCMs outputs in the northwestern and southwestern regions The black red blue and green lines represent the PDFs duringthe observed period (1984ndash2013) future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs) modelsrespectively for the period (2041ndash2070)

region the DH increases using the three GCMs results withranges between 203 and 291 192 and 278 and 199 and 273respectively Dinajpur subregion reveals the largest DH valueexcept future 3 GCMs model results where Rajshahi exhibitsthe highest DH values and Rangpur displays the smallestDH value in the northwestern region in the mid-centuryperiod On the other hand in the southwestern region theDH decreases almost half under three GCMs using the A1Bscenario compared to the current period with the rangesfrom 117 to 220 138 to 277 and 148 to 278 correspondinglyThe highest DH value was observed in Khulna subregionand the lowest DH value was detected in Faridpur in futureclimate condition The DH value showed a gentle increasingtrend in northern and western part of northwestern regionthan southern part of southwestern region for the period of2041 to 2070 Such findings are consistent with the previousstudy of Hossain et al [50] who revealed that future climatechange will increase both frequency and magnitude of severe

drought events in the northwestern region The results ofSelvaraju and Baas [34] are also in good agreement withthis outcomes at future climate change condition Underclimate change scenario for the 2050 they reported thatfuture drought will be increased in the northwestern tocentral region of Bangladesh The study demonstrates thatthe amplitude of DH variations will increase slightly inthe northwestern region and flatten with a decrease in thesouthwestern region under future climate change conditionsFrom the analysis results of three GCMs potential droughthazard areas will shift from the south toward the north andnorthwest parts of the country in future It is evident from thestudy that the high DH occurs in February during the paddygrowing season and low DH in April for the current periodbut it will shift under future climate change conditions thatthe high DHwill happen in January and lowDH in April It isnot surprising thatmore areas in the northwestern regionwillbe exposed to the severe drought hazard due to the projected

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

6 Advances in Meteorology

Table 3The results ofQ-test for precipitation and minimum and maximum temperature at 8 meteorological stations in the study area (unitnumber of station)

Variables P value Jan Feb Mar Apl May Jun Jul Aug Sep Oct Nov Dec

Precipitation lt005 1 0 0 0 0 2 1 0 1 0 3 1ge005 7 8 8 8 8 6 7 8 7 8 5 7

Minimum temperature lt005 0 1 0 0 0 0 0 0 0 0 0 0ge005 8 7 8 8 8 8 8 8 8 8 8 8

Maximum temperature lt005 0 0 0 0 0 0 0 0 0 0 0 0ge005 8 8 8 8 8 8 8 8 8 8 8 8

0 200 600Precipitation (mm)

ObservedSimulated

00000

00010

00020

Den

sity

minus200

(a)

ObservedSimulated

000

005

010

015

Den

sity

25 30 3520Max temp (∘C)

(b)

ObservedSimulated

000

002

004

006

Den

sity

10 20 305Min temp (∘C)

(c)

ObservedSimulated

00000

00010

00020

00030

Den

sity

0 200 400minus200

Precipitation (mm)

(d)

ObservedSimulated

000

005

010

015

Den

sity

30 35 4025Max temp (∘C)

(e)

ObservedSimulated

10 15 20 25 30 355000

002

004

006

008D

ensit

y

Min temp (∘C)

(f)

Figure 3 Comparison of observed and simulated climate variables in the paddy growing season (andashc) show precipitation and maximumand minimum temperature in Rangpur subregion northwestern Bangladesh (dndashf) represent precipitation and maximum and minimumtemperature in Satkhira subregion southwestern Bangladesh

monthly maximum and minimum temperature in case sta-tions (Figure 3) The justification is that these two represen-tative stations are used as an example because these stationsexhibit rainfall variability that are statistically significantTherainfall patterns of two stations are changing increasinglyand annual rainfall shows a declining trend [14] The modelshowed the better performance in the southwestern than thenorthwestern region for precipitation The linear regressionanalyses confirmed a very high correlation between theobserved and simulated precipitation (1199032 = 097 and 098)at 005 significance level in both regions Furthermorethe correlation coefficients (1199032) between the simulated and

observed temperature are around 099 The comparisonbetween simulated and observed data cannot guarantee theirvalidity of future climate because global warming plays a vitalrole in the near future climate change conditions which maynot be identical to the current climate

3 Result and Discussions

The ultimate goal of this study is to evaluate drought hazardin typical Boro paddy cultivated areas of western Bangladeshat present and future climate change conditions First wewill show the temporal variations in precipitation potential

Advances in Meteorology 7

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

RajshahiDinajpur Rangpur BograSubregions

50

100

150

200Pr

ecip

itatio

n (m

m)

(a) Northwestern region

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

Faridpur BarisalSatkhiraKhulnaSubregions

50

100

150

200

Prec

ipita

tion

(mm

)

(b) Southwestern region

Figure 4 The temporal variations in average precipitation during the paddy growing season under observed and future climate changeconditions in western Bangladesh (a) The changes of precipitation in the northwestern region (b) the changes of precipitation in thesouthwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 models respectively

evapotranspiration (PET) and drought events (SPEI lt minus1)under climate change during the Boro rice growing seasonin the designated areas This study illustrates the processesthat exaggerate drought hazard in the future Hereafterwe outline changes in precipitation PET and SPEI lt minus1from the observed (1984ndash2013) to future periods (2041ndash2070)in western Bangladesh The A1B scenario is used underthree GCMs outputs which are denoted as future model1 (CGCM31) future model 2 (FGOALS-G10) and futuremodel 3 (HadGEM1)

31 Changes of Precipitation PET and Drought Events underClimate Change

311 Changes in the Precipitation Figure 4 presents the tem-poral variation in average precipitation during the growingseason (January toMay) for the current and future periods inthemajor Boro paddy cultivated areas in western BangladeshAll the subregions showed that the average precipitation dur-ing the Boro rice growing season in northwestern Bangladeshwere 9495mm (future model 1) 7877mm (future model2) and 8163mm (future model 3) by the period of 2041ndash2070 compared to 6720mm in the current period of 1984ndash2013The highest average seasonal precipitationwas observedin Rangpur (14091mm) using future model 1 while thelowest seasonal precipitation was found in Bogra (5688mm)under future model 2 However Dinajpur demonstrated adecrease in precipitation by the period of 2041ndash2070 underfuture model 2 and future model 3 compared to the currentperiod Conversely in the case of the southwestern regionall subregions exhibited that average seasonal precipitationwas 9447mm 7653mm and 7226mm respectively in nearfuture period and 7218mm in the baseline periodThe largestaverage seasonal precipitation was in Faridpur (11218mm)under future model 1 while the smallest average seasonalprecipitation was in Satkhira (5859mm) using future model3 (Figure 4) From the results it is clear that the temporalvariations in precipitation are higher in the northwesternregion than the southwestern Bangladesh The increasingtendency of precipitation under potential climate changesmay be due to high variations in rainy days amounts ofprecipitation and intensity in the Boro rice growing season in

Bangladesh In addition the huge water vapor transport fromthe Bay of Bengal and enhanced moisture convergence maybe another cause of increasing precipitation in future [33]

312 Changes in the Potential Evapotranspiration Figure 5exhibits future change in average PET in the Boro ricegrowing season for major subregions of western Bangladeshat the current and future periods under A1B scenario Theaverage seasonal PET during January to May in the north-western region was 11141mm in the observed period andincrease in the near future about 14763mm (future model 1)13487mm (future model 2) and 14542mm (future model3) respectively by the period of 2041 to 2070 The highestaverage seasonal PET was detected in Rajshahi (17035mm)under future model 3 whereas the lowest average PET wasobserved in Rangpur (12041mm) using future model 2(Figure 5)

On the other hand average PET of four subregions inthe southwestern Bangladesh was 13327mm for the currentperiod and increase in 17389 15922mm and 18304mmrespectively for future period using the three GCMs resultsThe largest seasonal average PET was noticed in Satkhira(19511mm) under future model 3 while the smallest PETwas observed in Faridpur (15043mm) using future model 2(Figure 5)The results show that the PET variations in north-western region are much larger compared to that in south-western region Furthermore the discrepancies between thePET changes using the three GCMs become more noticeablein the mid-century period This finding is consistent withthe previous results of Shahid [6] that showed an increaserate of potential evapotranspiration (PET) in northwesternBangladesh The widespread increase in PET is expected torise in temperature and amounts of solar radiation underclimate change conditions that dominate in the western partof Bangladesh Although the tendency of precipitation andPET are both upward trends the increasing rate of PETexceeds that of precipitation indicating that the increasingprecipitation is not offsetting effects of higher evaporationThe rise in global temperature and surface warming playsa dominant role in the future PET enhancement The studyimplies that enhancement of PET in the study area willaggravate drought frequency and intensity in the near future

8 Advances in Meteorology

Dinajpur RangpurSubregions

Bogra Rajshahi

100120140160180

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(a) Northwestern region

Faridpur KhulnaSubregions

BarisalSatkhira

100120140160180200

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(b) Southwestern region

Figure 5The temporal variations in average PET during the growing season under observed and future climate change conditions in westernBangladesh (a) The changes of potential evapotranspiration (PET) in northwestern region (b) the changes of potential evapotranspiration(PET) in southwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 modelsrespectively

313 Changes in the Drought Events Figure 6 shows theprobability density function (PDF) of the SPEI by thethreshold level during the Boro paddy cultivating season innorthwestern and southwestern regions under current andfuture climate change conditions at monthly time scale Thetotal SPEI valuesltminus1 during January toMay at current period(1984ndash2013) are mainly normally distributed according to themathematical principle of the SPEI procedure The resultsreveal that the PDFs become broader with future time andpeak probability distribution will increase in the near futurefor the period between 2041 and 2070 in comparison to theobserved period For instance Dinajpur subregion shows asevere increased drought events in terms of intensity and fre-quency under A1B scenario using future model 3 (HadGEM1GCMs) In addition Rajsahhi and Bogra subregions exhibitmore drought intensity in potential future climate changecompared to the current period The results show that thePDFs will move the upper tail continuously toward the mid-century period and also shift to some extent in left side(Figure 6)

Figure 6 exhibits the trend of a higher increase in the SPEIltminus1 in Khulna and Satkhira in future period compared to theothers subregions in the southwestern Bangladesh Compar-ing to drought events in the southwestern region with thenorthwestern region it is obvious that the PDFs of monthlytotal SPEI values below minus1 in the Boro paddy cultivatedarea will change drastically in the near future The futuremodel 1 (CGCM31 GCMs) and future model 2 (FGOALS-G10 GCMs) indicate that frequency of extreme droughtevents (SPEI lt minus1) is expected to increase dramatically in thewestern Bangladesh It is worth mentioning that the risk ofextreme drought events tends to increase in future climateconditions Besides drought frequency and intensity implya steady increase from the current to mid-century periodsbecause the PDFs of SPEI values shift more than one and ahalf standard deviation peaking at minus15 in all the subregionsof western Bangladesh during that periodThe results of threeGCMs show that the upper tail of PDFswill become thicker infuture period (Figure 6)The study indicates that the PDFs ofdrought events will continue tomove upward in future periodcompared to the current period The results demonstrate

that the northwestern region will increase a severe droughtfrequency and levels of intensity in comparison with thesouthwestern region in the near future

32 Spatial Variation in Drought Hazard under ClimateChange The spatial variation in drought hazard (DH) foreach subregion in western Bangladesh is computed basedon the LARS-WG downscaled precipitation and temperaturedata for the period between 2041 and 2070 Figure 7 displaysthe spatial distribution in DH during the Boro rice growingseason using the results of future 1 (CGCM31 GCMs) future2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs)models respectively

The DH value in the northwestern region ranges from180 to 261 for the current period with the highest DHvalue distributed in Rajshahi subregion and the lowest DHvalue exhibited in Rangpur subregion Conversely the DHvalue in the southwestern region ranges between 239 and376 during the baseline period (1984ndash2013) The highest DHvalue was found in Satkhira subregion while the lowest DHvalue was distributed in Faridpur subregion The obtainedresults indicated that the amplitude of DH was much largerin the southwestern region than the northwestern regionduring the period of 1984 to 2013 The findings are consistentwith the earlier studies of Abedin et al [48] and Abdullah[15] they reported that southwestern coastal region is muchdrought-prone in comparison to the rest of the country Butthe findings of Shahid and Behrawan [14] Alamgir et al [16]and Islam et al [49] disagree with the present observationof drought hazard analysis They found that drought poseda high risk to northern and northwestern part than southernand southwestern parts of Bangladesh in the current periodShahid and Behrawan [14] used onlymonthly rainfall data forthe period of 1961ndash1999 with old datasets in the western partof Bangladesh and did not consider temperature data Thepresent study has taken into consideration both precipitationand temperature data (1984ndash2070) under current and futureclimate change conditions in the winter Boro paddy growingseason of western Bangladesh This is the reason that thefindings are not consistent with those previous studies In thenear future period (2041ndash2070) especially the northwestern

Advances in Meteorology 9

ObservedFuture 1

Future 2Future 3

0 2minus2minus4

SPEI

00

02

04

06

Den

sity

(a) Dinajpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

12

Den

sity

(b) Rangpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

Den

sity

(c) Bogra

ObservedFuture 1

Future 2Future 3

0 1minus4 minus2minus6

SPEI

00

04

Den

sity

(d) Rajshahi

ObservedFuture 1

Future 2Future 3

minus4 0minus6 minus2

SPEI

00

02

04

06

Den

sity

(e) Faridpur

ObservedFuture 1

Future 2Future 3

minus4 minus2 0 2minus6

SPEI

00

02

04

06

Den

sity

(f) Khulna

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

03

06

Den

sity

(g) Satkhira

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

02

04

Den

sity

(h) Barisal

Figure 6 The probability density function (PDF) of SPEI values by the threshold level during January to May for the A1B scenario usingthe three GCMs outputs in the northwestern and southwestern regions The black red blue and green lines represent the PDFs duringthe observed period (1984ndash2013) future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs) modelsrespectively for the period (2041ndash2070)

region the DH increases using the three GCMs results withranges between 203 and 291 192 and 278 and 199 and 273respectively Dinajpur subregion reveals the largest DH valueexcept future 3 GCMs model results where Rajshahi exhibitsthe highest DH values and Rangpur displays the smallestDH value in the northwestern region in the mid-centuryperiod On the other hand in the southwestern region theDH decreases almost half under three GCMs using the A1Bscenario compared to the current period with the rangesfrom 117 to 220 138 to 277 and 148 to 278 correspondinglyThe highest DH value was observed in Khulna subregionand the lowest DH value was detected in Faridpur in futureclimate condition The DH value showed a gentle increasingtrend in northern and western part of northwestern regionthan southern part of southwestern region for the period of2041 to 2070 Such findings are consistent with the previousstudy of Hossain et al [50] who revealed that future climatechange will increase both frequency and magnitude of severe

drought events in the northwestern region The results ofSelvaraju and Baas [34] are also in good agreement withthis outcomes at future climate change condition Underclimate change scenario for the 2050 they reported thatfuture drought will be increased in the northwestern tocentral region of Bangladesh The study demonstrates thatthe amplitude of DH variations will increase slightly inthe northwestern region and flatten with a decrease in thesouthwestern region under future climate change conditionsFrom the analysis results of three GCMs potential droughthazard areas will shift from the south toward the north andnorthwest parts of the country in future It is evident from thestudy that the high DH occurs in February during the paddygrowing season and low DH in April for the current periodbut it will shift under future climate change conditions thatthe high DHwill happen in January and lowDH in April It isnot surprising thatmore areas in the northwestern regionwillbe exposed to the severe drought hazard due to the projected

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

Advances in Meteorology 7

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

RajshahiDinajpur Rangpur BograSubregions

50

100

150

200Pr

ecip

itatio

n (m

m)

(a) Northwestern region

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

Faridpur BarisalSatkhiraKhulnaSubregions

50

100

150

200

Prec

ipita

tion

(mm

)

(b) Southwestern region

Figure 4 The temporal variations in average precipitation during the paddy growing season under observed and future climate changeconditions in western Bangladesh (a) The changes of precipitation in the northwestern region (b) the changes of precipitation in thesouthwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 models respectively

evapotranspiration (PET) and drought events (SPEI lt minus1)under climate change during the Boro rice growing seasonin the designated areas This study illustrates the processesthat exaggerate drought hazard in the future Hereafterwe outline changes in precipitation PET and SPEI lt minus1from the observed (1984ndash2013) to future periods (2041ndash2070)in western Bangladesh The A1B scenario is used underthree GCMs outputs which are denoted as future model1 (CGCM31) future model 2 (FGOALS-G10) and futuremodel 3 (HadGEM1)

31 Changes of Precipitation PET and Drought Events underClimate Change

311 Changes in the Precipitation Figure 4 presents the tem-poral variation in average precipitation during the growingseason (January toMay) for the current and future periods inthemajor Boro paddy cultivated areas in western BangladeshAll the subregions showed that the average precipitation dur-ing the Boro rice growing season in northwestern Bangladeshwere 9495mm (future model 1) 7877mm (future model2) and 8163mm (future model 3) by the period of 2041ndash2070 compared to 6720mm in the current period of 1984ndash2013The highest average seasonal precipitationwas observedin Rangpur (14091mm) using future model 1 while thelowest seasonal precipitation was found in Bogra (5688mm)under future model 2 However Dinajpur demonstrated adecrease in precipitation by the period of 2041ndash2070 underfuture model 2 and future model 3 compared to the currentperiod Conversely in the case of the southwestern regionall subregions exhibited that average seasonal precipitationwas 9447mm 7653mm and 7226mm respectively in nearfuture period and 7218mm in the baseline periodThe largestaverage seasonal precipitation was in Faridpur (11218mm)under future model 1 while the smallest average seasonalprecipitation was in Satkhira (5859mm) using future model3 (Figure 4) From the results it is clear that the temporalvariations in precipitation are higher in the northwesternregion than the southwestern Bangladesh The increasingtendency of precipitation under potential climate changesmay be due to high variations in rainy days amounts ofprecipitation and intensity in the Boro rice growing season in

Bangladesh In addition the huge water vapor transport fromthe Bay of Bengal and enhanced moisture convergence maybe another cause of increasing precipitation in future [33]

312 Changes in the Potential Evapotranspiration Figure 5exhibits future change in average PET in the Boro ricegrowing season for major subregions of western Bangladeshat the current and future periods under A1B scenario Theaverage seasonal PET during January to May in the north-western region was 11141mm in the observed period andincrease in the near future about 14763mm (future model 1)13487mm (future model 2) and 14542mm (future model3) respectively by the period of 2041 to 2070 The highestaverage seasonal PET was detected in Rajshahi (17035mm)under future model 3 whereas the lowest average PET wasobserved in Rangpur (12041mm) using future model 2(Figure 5)

On the other hand average PET of four subregions inthe southwestern Bangladesh was 13327mm for the currentperiod and increase in 17389 15922mm and 18304mmrespectively for future period using the three GCMs resultsThe largest seasonal average PET was noticed in Satkhira(19511mm) under future model 3 while the smallest PETwas observed in Faridpur (15043mm) using future model 2(Figure 5)The results show that the PET variations in north-western region are much larger compared to that in south-western region Furthermore the discrepancies between thePET changes using the three GCMs become more noticeablein the mid-century period This finding is consistent withthe previous results of Shahid [6] that showed an increaserate of potential evapotranspiration (PET) in northwesternBangladesh The widespread increase in PET is expected torise in temperature and amounts of solar radiation underclimate change conditions that dominate in the western partof Bangladesh Although the tendency of precipitation andPET are both upward trends the increasing rate of PETexceeds that of precipitation indicating that the increasingprecipitation is not offsetting effects of higher evaporationThe rise in global temperature and surface warming playsa dominant role in the future PET enhancement The studyimplies that enhancement of PET in the study area willaggravate drought frequency and intensity in the near future

8 Advances in Meteorology

Dinajpur RangpurSubregions

Bogra Rajshahi

100120140160180

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(a) Northwestern region

Faridpur KhulnaSubregions

BarisalSatkhira

100120140160180200

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(b) Southwestern region

Figure 5The temporal variations in average PET during the growing season under observed and future climate change conditions in westernBangladesh (a) The changes of potential evapotranspiration (PET) in northwestern region (b) the changes of potential evapotranspiration(PET) in southwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 modelsrespectively

313 Changes in the Drought Events Figure 6 shows theprobability density function (PDF) of the SPEI by thethreshold level during the Boro paddy cultivating season innorthwestern and southwestern regions under current andfuture climate change conditions at monthly time scale Thetotal SPEI valuesltminus1 during January toMay at current period(1984ndash2013) are mainly normally distributed according to themathematical principle of the SPEI procedure The resultsreveal that the PDFs become broader with future time andpeak probability distribution will increase in the near futurefor the period between 2041 and 2070 in comparison to theobserved period For instance Dinajpur subregion shows asevere increased drought events in terms of intensity and fre-quency under A1B scenario using future model 3 (HadGEM1GCMs) In addition Rajsahhi and Bogra subregions exhibitmore drought intensity in potential future climate changecompared to the current period The results show that thePDFs will move the upper tail continuously toward the mid-century period and also shift to some extent in left side(Figure 6)

Figure 6 exhibits the trend of a higher increase in the SPEIltminus1 in Khulna and Satkhira in future period compared to theothers subregions in the southwestern Bangladesh Compar-ing to drought events in the southwestern region with thenorthwestern region it is obvious that the PDFs of monthlytotal SPEI values below minus1 in the Boro paddy cultivatedarea will change drastically in the near future The futuremodel 1 (CGCM31 GCMs) and future model 2 (FGOALS-G10 GCMs) indicate that frequency of extreme droughtevents (SPEI lt minus1) is expected to increase dramatically in thewestern Bangladesh It is worth mentioning that the risk ofextreme drought events tends to increase in future climateconditions Besides drought frequency and intensity implya steady increase from the current to mid-century periodsbecause the PDFs of SPEI values shift more than one and ahalf standard deviation peaking at minus15 in all the subregionsof western Bangladesh during that periodThe results of threeGCMs show that the upper tail of PDFswill become thicker infuture period (Figure 6)The study indicates that the PDFs ofdrought events will continue tomove upward in future periodcompared to the current period The results demonstrate

that the northwestern region will increase a severe droughtfrequency and levels of intensity in comparison with thesouthwestern region in the near future

32 Spatial Variation in Drought Hazard under ClimateChange The spatial variation in drought hazard (DH) foreach subregion in western Bangladesh is computed basedon the LARS-WG downscaled precipitation and temperaturedata for the period between 2041 and 2070 Figure 7 displaysthe spatial distribution in DH during the Boro rice growingseason using the results of future 1 (CGCM31 GCMs) future2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs)models respectively

The DH value in the northwestern region ranges from180 to 261 for the current period with the highest DHvalue distributed in Rajshahi subregion and the lowest DHvalue exhibited in Rangpur subregion Conversely the DHvalue in the southwestern region ranges between 239 and376 during the baseline period (1984ndash2013) The highest DHvalue was found in Satkhira subregion while the lowest DHvalue was distributed in Faridpur subregion The obtainedresults indicated that the amplitude of DH was much largerin the southwestern region than the northwestern regionduring the period of 1984 to 2013 The findings are consistentwith the earlier studies of Abedin et al [48] and Abdullah[15] they reported that southwestern coastal region is muchdrought-prone in comparison to the rest of the country Butthe findings of Shahid and Behrawan [14] Alamgir et al [16]and Islam et al [49] disagree with the present observationof drought hazard analysis They found that drought poseda high risk to northern and northwestern part than southernand southwestern parts of Bangladesh in the current periodShahid and Behrawan [14] used onlymonthly rainfall data forthe period of 1961ndash1999 with old datasets in the western partof Bangladesh and did not consider temperature data Thepresent study has taken into consideration both precipitationand temperature data (1984ndash2070) under current and futureclimate change conditions in the winter Boro paddy growingseason of western Bangladesh This is the reason that thefindings are not consistent with those previous studies In thenear future period (2041ndash2070) especially the northwestern

Advances in Meteorology 9

ObservedFuture 1

Future 2Future 3

0 2minus2minus4

SPEI

00

02

04

06

Den

sity

(a) Dinajpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

12

Den

sity

(b) Rangpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

Den

sity

(c) Bogra

ObservedFuture 1

Future 2Future 3

0 1minus4 minus2minus6

SPEI

00

04

Den

sity

(d) Rajshahi

ObservedFuture 1

Future 2Future 3

minus4 0minus6 minus2

SPEI

00

02

04

06

Den

sity

(e) Faridpur

ObservedFuture 1

Future 2Future 3

minus4 minus2 0 2minus6

SPEI

00

02

04

06

Den

sity

(f) Khulna

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

03

06

Den

sity

(g) Satkhira

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

02

04

Den

sity

(h) Barisal

Figure 6 The probability density function (PDF) of SPEI values by the threshold level during January to May for the A1B scenario usingthe three GCMs outputs in the northwestern and southwestern regions The black red blue and green lines represent the PDFs duringthe observed period (1984ndash2013) future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs) modelsrespectively for the period (2041ndash2070)

region the DH increases using the three GCMs results withranges between 203 and 291 192 and 278 and 199 and 273respectively Dinajpur subregion reveals the largest DH valueexcept future 3 GCMs model results where Rajshahi exhibitsthe highest DH values and Rangpur displays the smallestDH value in the northwestern region in the mid-centuryperiod On the other hand in the southwestern region theDH decreases almost half under three GCMs using the A1Bscenario compared to the current period with the rangesfrom 117 to 220 138 to 277 and 148 to 278 correspondinglyThe highest DH value was observed in Khulna subregionand the lowest DH value was detected in Faridpur in futureclimate condition The DH value showed a gentle increasingtrend in northern and western part of northwestern regionthan southern part of southwestern region for the period of2041 to 2070 Such findings are consistent with the previousstudy of Hossain et al [50] who revealed that future climatechange will increase both frequency and magnitude of severe

drought events in the northwestern region The results ofSelvaraju and Baas [34] are also in good agreement withthis outcomes at future climate change condition Underclimate change scenario for the 2050 they reported thatfuture drought will be increased in the northwestern tocentral region of Bangladesh The study demonstrates thatthe amplitude of DH variations will increase slightly inthe northwestern region and flatten with a decrease in thesouthwestern region under future climate change conditionsFrom the analysis results of three GCMs potential droughthazard areas will shift from the south toward the north andnorthwest parts of the country in future It is evident from thestudy that the high DH occurs in February during the paddygrowing season and low DH in April for the current periodbut it will shift under future climate change conditions thatthe high DHwill happen in January and lowDH in April It isnot surprising thatmore areas in the northwestern regionwillbe exposed to the severe drought hazard due to the projected

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

8 Advances in Meteorology

Dinajpur RangpurSubregions

Bogra Rajshahi

100120140160180

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(a) Northwestern region

Faridpur KhulnaSubregions

BarisalSatkhira

100120140160180200

PET

(mm

)

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

OBS

FM-1

FM-2

FM-3

(b) Southwestern region

Figure 5The temporal variations in average PET during the growing season under observed and future climate change conditions in westernBangladesh (a) The changes of potential evapotranspiration (PET) in northwestern region (b) the changes of potential evapotranspiration(PET) in southwestern region FM-1 FM-2 and FM-3 indicate three GCMs namely CGCM31 FGOALS-G10 and HadGEM1 modelsrespectively

313 Changes in the Drought Events Figure 6 shows theprobability density function (PDF) of the SPEI by thethreshold level during the Boro paddy cultivating season innorthwestern and southwestern regions under current andfuture climate change conditions at monthly time scale Thetotal SPEI valuesltminus1 during January toMay at current period(1984ndash2013) are mainly normally distributed according to themathematical principle of the SPEI procedure The resultsreveal that the PDFs become broader with future time andpeak probability distribution will increase in the near futurefor the period between 2041 and 2070 in comparison to theobserved period For instance Dinajpur subregion shows asevere increased drought events in terms of intensity and fre-quency under A1B scenario using future model 3 (HadGEM1GCMs) In addition Rajsahhi and Bogra subregions exhibitmore drought intensity in potential future climate changecompared to the current period The results show that thePDFs will move the upper tail continuously toward the mid-century period and also shift to some extent in left side(Figure 6)

Figure 6 exhibits the trend of a higher increase in the SPEIltminus1 in Khulna and Satkhira in future period compared to theothers subregions in the southwestern Bangladesh Compar-ing to drought events in the southwestern region with thenorthwestern region it is obvious that the PDFs of monthlytotal SPEI values below minus1 in the Boro paddy cultivatedarea will change drastically in the near future The futuremodel 1 (CGCM31 GCMs) and future model 2 (FGOALS-G10 GCMs) indicate that frequency of extreme droughtevents (SPEI lt minus1) is expected to increase dramatically in thewestern Bangladesh It is worth mentioning that the risk ofextreme drought events tends to increase in future climateconditions Besides drought frequency and intensity implya steady increase from the current to mid-century periodsbecause the PDFs of SPEI values shift more than one and ahalf standard deviation peaking at minus15 in all the subregionsof western Bangladesh during that periodThe results of threeGCMs show that the upper tail of PDFswill become thicker infuture period (Figure 6)The study indicates that the PDFs ofdrought events will continue tomove upward in future periodcompared to the current period The results demonstrate

that the northwestern region will increase a severe droughtfrequency and levels of intensity in comparison with thesouthwestern region in the near future

32 Spatial Variation in Drought Hazard under ClimateChange The spatial variation in drought hazard (DH) foreach subregion in western Bangladesh is computed basedon the LARS-WG downscaled precipitation and temperaturedata for the period between 2041 and 2070 Figure 7 displaysthe spatial distribution in DH during the Boro rice growingseason using the results of future 1 (CGCM31 GCMs) future2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs)models respectively

The DH value in the northwestern region ranges from180 to 261 for the current period with the highest DHvalue distributed in Rajshahi subregion and the lowest DHvalue exhibited in Rangpur subregion Conversely the DHvalue in the southwestern region ranges between 239 and376 during the baseline period (1984ndash2013) The highest DHvalue was found in Satkhira subregion while the lowest DHvalue was distributed in Faridpur subregion The obtainedresults indicated that the amplitude of DH was much largerin the southwestern region than the northwestern regionduring the period of 1984 to 2013 The findings are consistentwith the earlier studies of Abedin et al [48] and Abdullah[15] they reported that southwestern coastal region is muchdrought-prone in comparison to the rest of the country Butthe findings of Shahid and Behrawan [14] Alamgir et al [16]and Islam et al [49] disagree with the present observationof drought hazard analysis They found that drought poseda high risk to northern and northwestern part than southernand southwestern parts of Bangladesh in the current periodShahid and Behrawan [14] used onlymonthly rainfall data forthe period of 1961ndash1999 with old datasets in the western partof Bangladesh and did not consider temperature data Thepresent study has taken into consideration both precipitationand temperature data (1984ndash2070) under current and futureclimate change conditions in the winter Boro paddy growingseason of western Bangladesh This is the reason that thefindings are not consistent with those previous studies In thenear future period (2041ndash2070) especially the northwestern

Advances in Meteorology 9

ObservedFuture 1

Future 2Future 3

0 2minus2minus4

SPEI

00

02

04

06

Den

sity

(a) Dinajpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

12

Den

sity

(b) Rangpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

Den

sity

(c) Bogra

ObservedFuture 1

Future 2Future 3

0 1minus4 minus2minus6

SPEI

00

04

Den

sity

(d) Rajshahi

ObservedFuture 1

Future 2Future 3

minus4 0minus6 minus2

SPEI

00

02

04

06

Den

sity

(e) Faridpur

ObservedFuture 1

Future 2Future 3

minus4 minus2 0 2minus6

SPEI

00

02

04

06

Den

sity

(f) Khulna

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

03

06

Den

sity

(g) Satkhira

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

02

04

Den

sity

(h) Barisal

Figure 6 The probability density function (PDF) of SPEI values by the threshold level during January to May for the A1B scenario usingthe three GCMs outputs in the northwestern and southwestern regions The black red blue and green lines represent the PDFs duringthe observed period (1984ndash2013) future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs) modelsrespectively for the period (2041ndash2070)

region the DH increases using the three GCMs results withranges between 203 and 291 192 and 278 and 199 and 273respectively Dinajpur subregion reveals the largest DH valueexcept future 3 GCMs model results where Rajshahi exhibitsthe highest DH values and Rangpur displays the smallestDH value in the northwestern region in the mid-centuryperiod On the other hand in the southwestern region theDH decreases almost half under three GCMs using the A1Bscenario compared to the current period with the rangesfrom 117 to 220 138 to 277 and 148 to 278 correspondinglyThe highest DH value was observed in Khulna subregionand the lowest DH value was detected in Faridpur in futureclimate condition The DH value showed a gentle increasingtrend in northern and western part of northwestern regionthan southern part of southwestern region for the period of2041 to 2070 Such findings are consistent with the previousstudy of Hossain et al [50] who revealed that future climatechange will increase both frequency and magnitude of severe

drought events in the northwestern region The results ofSelvaraju and Baas [34] are also in good agreement withthis outcomes at future climate change condition Underclimate change scenario for the 2050 they reported thatfuture drought will be increased in the northwestern tocentral region of Bangladesh The study demonstrates thatthe amplitude of DH variations will increase slightly inthe northwestern region and flatten with a decrease in thesouthwestern region under future climate change conditionsFrom the analysis results of three GCMs potential droughthazard areas will shift from the south toward the north andnorthwest parts of the country in future It is evident from thestudy that the high DH occurs in February during the paddygrowing season and low DH in April for the current periodbut it will shift under future climate change conditions thatthe high DHwill happen in January and lowDH in April It isnot surprising thatmore areas in the northwestern regionwillbe exposed to the severe drought hazard due to the projected

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

Advances in Meteorology 9

ObservedFuture 1

Future 2Future 3

0 2minus2minus4

SPEI

00

02

04

06

Den

sity

(a) Dinajpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

12

Den

sity

(b) Rangpur

ObservedFuture 1

Future 2Future 3

minus5 minus1 0 1minus3

SPEI

00

04

08

Den

sity

(c) Bogra

ObservedFuture 1

Future 2Future 3

0 1minus4 minus2minus6

SPEI

00

04

Den

sity

(d) Rajshahi

ObservedFuture 1

Future 2Future 3

minus4 0minus6 minus2

SPEI

00

02

04

06

Den

sity

(e) Faridpur

ObservedFuture 1

Future 2Future 3

minus4 minus2 0 2minus6

SPEI

00

02

04

06

Den

sity

(f) Khulna

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

03

06

Den

sity

(g) Satkhira

ObservedFuture 1

Future 2Future 3

minus6 minus4 minus2 20SPEI

00

02

04

Den

sity

(h) Barisal

Figure 6 The probability density function (PDF) of SPEI values by the threshold level during January to May for the A1B scenario usingthe three GCMs outputs in the northwestern and southwestern regions The black red blue and green lines represent the PDFs duringthe observed period (1984ndash2013) future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs) and future 3 (HadGEM1 GCMs) modelsrespectively for the period (2041ndash2070)

region the DH increases using the three GCMs results withranges between 203 and 291 192 and 278 and 199 and 273respectively Dinajpur subregion reveals the largest DH valueexcept future 3 GCMs model results where Rajshahi exhibitsthe highest DH values and Rangpur displays the smallestDH value in the northwestern region in the mid-centuryperiod On the other hand in the southwestern region theDH decreases almost half under three GCMs using the A1Bscenario compared to the current period with the rangesfrom 117 to 220 138 to 277 and 148 to 278 correspondinglyThe highest DH value was observed in Khulna subregionand the lowest DH value was detected in Faridpur in futureclimate condition The DH value showed a gentle increasingtrend in northern and western part of northwestern regionthan southern part of southwestern region for the period of2041 to 2070 Such findings are consistent with the previousstudy of Hossain et al [50] who revealed that future climatechange will increase both frequency and magnitude of severe

drought events in the northwestern region The results ofSelvaraju and Baas [34] are also in good agreement withthis outcomes at future climate change condition Underclimate change scenario for the 2050 they reported thatfuture drought will be increased in the northwestern tocentral region of Bangladesh The study demonstrates thatthe amplitude of DH variations will increase slightly inthe northwestern region and flatten with a decrease in thesouthwestern region under future climate change conditionsFrom the analysis results of three GCMs potential droughthazard areas will shift from the south toward the north andnorthwest parts of the country in future It is evident from thestudy that the high DH occurs in February during the paddygrowing season and low DH in April for the current periodbut it will shift under future climate change conditions thatthe high DHwill happen in January and lowDH in April It isnot surprising thatmore areas in the northwestern regionwillbe exposed to the severe drought hazard due to the projected

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

10 Advances in Meteorology

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Bogra

Rangpur

Rajshahi

Dinajpur

Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira Khulna Barisal

Faridpur

SatkhiraKhulna Barisal

Faridpur

Satkhira

Bogra

Rangpur

Rajshahi

Dinajpur

Northwestern region

Southwestern region

Observed Future 3Future 1 Future 2

0 75 150 225 300375(km) Drought hazard value

N

15

ndash2

21

ndash25

31

ndash35

251

ndash3

351

ndash4

Figure 7The spatial distributionmaps show drought hazard during the Boro paddy growing season of western Bangladesh for future period(2041ndash2070) relative to the observed period (1984ndash2013) Three GCMs denote future 1 (CGCM31 GCMs) future 2 (FGOALS-G10 GCMs)and future 3 (HadGEM1 GCMs) model respectively

change in rainfall patterns and PET frequencies with futureclimate change condition

4 Conclusions

This study presents the drought hazard index in terms ofdrought frequency and intensity to evaluate the Boro paddygrowing season drought hazard in western Bangladesh Wesimulated a climate database for future periods (2041ndash2070)by the downscaled LARS-WG model using the outputs ofthree GCMs namely CGCM31 (Canada) FGOALS-G10(China) and HadGEM1 (UK) based on the SRES A1Bscenario The study provides a comprehensive idea about thefrequency and intensity of drought events during the Bororice growing season at present and future climate changeconditions The results show that changes in precipitationare fairly varied with some decrease in the northwesternregion and an increase in the southwestern region in futureperiod between 2041 and 2070 The overall increase in PETis closely related to rises in temperature and surface netradiation that dominates in the both regions of BangladeshThe PET and precipitation exhibit upward trends implyingthat the increased rate of precipitation is not outweighed thePET In this study the SPEI is used to quantify future droughtchanges and the increases in PET with surface warming areanticipated to make more drought events in future Fromthe results of three GCMs drought hazard will marginally

increase in the northwestern region and a decrease inthe southwestern region in the mid-century period Themost important finding is that high drought hazards in thenorthwestern region are mainly due to the increased rate ofPET exceeding that of precipitation in changing climate Inaddition potential drought hazard region will shift from thesouthwest to northwest parts of the country in the near futureas projected by using the three GCMs

It is noted that the results of three GCMs using thesimulated precipitation and temperature data and uncertain-ties of the downscaled LARS-WG model to employ a singlescenario may have caused some setback in the accuracy ofour results Future studies concentrating on the following willbe necessary comparing different GCMs results in multiplescenarios improving downscaled method assessing droughtrisk from the drought hazard combined with vulnerabilityand exposure to the wholeBoro cultivated area of the countrySince the projected change of rainfall patterns and PETfrequencies under climate change condition particularly incontext of global warming having a severe impact on droughtit is essential to formulate the suitable drought adaptationpolicies and effectively implement these policies with a betterprovision from the government and nongovernment organi-zations (NGOs) It can be said that priority should be givenin the northwestern region compared to the southwesternregion for making future drought management strategiesof Bangladesh As this study evaluated the drought hazard

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

Advances in Meteorology 11

in the Boro cultivated areas it is anticipated that this willbe a helpful guide to understanding drought events andaid in formulating a broad adaptation strategy under futureclimate change condition to overcome the drought problemsuccessfully in the western Bangladesh

Disclosure

Abu Reza Md Towfiqul Islam is the first author

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

AbuRezaMd Towfiqul Islam thanks theChinese ScholarshipCouncil (CSC) for the financial support and the NanjingUniversity of Information Science and Technology (NUIST)and the Begum Rokeya University Rangpur Bangladeshfor different forms of support during the study period Thisstudy was jointly supported by the China Special Fundfor Meteorological Research in the Public Interest (MajorProject GYHY201506001-6) and the Special Fund of Mete-orological Scientific Research for the PublicWelfare of China(GYHY 201506018) under ldquoResponse Mechanism of Rice inHigh Temperature and Coping Technologyrdquo The authorsalso acknowledge Bangladesh Meteorological Department(BMD) for providing historical climate data used in thisstudy

References

[1] A Ali ldquoVulnerability of Bangladesh to climate change and sealevel rise through tropical cyclones and storm surgesrdquo WaterAir and Soil Pollution vol 92 no 1-2 pp 171ndash179 1996

[2] M A Abedin and R Shaw Climate Change Adaptation Actionsin Bangladesh Disaster Risk Reduction Springer Tokyo Japan2013

[3] C Rosenzweig J Elliott D Deryng et al ldquoAssessing agri-cultural risks of climate change in the 21st century in aglobal gridded crop model intercomparisonrdquo Proceedings of theNational Academy of Sciences vol 111 no 9 pp 3268ndash3273 2014

[4] IPCC Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge UK 2007

[5] R Selvaraju A R Subbiah S Baas and I Juergens LivelihoodAdaptation to Climate Variability and Change in Drought-Prone Areas of Bangladesh Developing Institutions and OptionsAsian Disaster Preparedness Centre Food and AgricultureOrganization Rome Italy 2006

[6] S Shahid ldquoImpact of climate change on irrigation waterdemand of dry season Boro rice in northwest BangladeshrdquoClimatic Change vol 105 no 3-4 pp 433ndash453 2011

[7] National DroughtMitigation Center ldquoWhat is drought Under-standing and defining droughtrdquo 2008 httpwwwdroughtunledu

[8] T B McKee N J Doesken and J Kleist ldquoThe relationship ofdrought frequency and duration to time scalesrdquo in Proceedingsof the 8th Conference on Applied Climatology pp 179ndash183Anaheim Calif USA January 1993

[9] W C Palmer ldquoMeteorological droughtrdquo Research Paper 45Department of Commerce Washington DC USA 1965

[10] S M Vicente-Serrano S Beguerıa and J I Lopez-MorenoldquoA multiscalar drought index sensitive to global warming thestandardized precipitation evapotranspiration indexrdquo Journal ofClimate vol 23 no 7 pp 1696ndash1718 2010

[11] M J Hayes M D Svoboda D AWilhite andO V VanyarkholdquoMonitoring the 1996 drought using the Standardized Precipita-tion Indexrdquo Bulletin of the American Meteorological Society vol80 no 3 pp 429ndash438 1999

[12] Z Wang J Jiang Y Liao and L Deng ldquoRisk assessment ofmaize drought hazard in the middle region of farming-pastoralecotone in Northern ChinardquoNatural Hazards vol 76 no 3 pp1515ndash1534 2015

[13] M Dubrovsky M D Svoboda M Trnka et al ldquoApplicationof relative drought indices in assessing climate-change impactson drought conditions in Czechiardquo Theoretical and AppliedClimatology vol 96 no 1-2 pp 155ndash171 2009

[14] S Shahid and H Behrawan ldquoDrought risk assessment in thewestern part of BangladeshrdquoNatural Hazards vol 46 no 3 pp391ndash413 2008

[15] S M Abdullah ldquoStandardized precipitation evapotranspirationindex (SPEI) based drought assessment in Bangladeshrdquo inProceedings of the 5th Proceedings of International Conferenceon Environmental Aspects of Bangladesh (ICEAB rsquo14) pp 40ndash44September 2014

[16] M Alamgir S Shahid M K Hazarika S Nashrrullah S BHarun and S Shamsudin ldquoAnalysis of meteorological droughtpattern during different climatic and cropping seasons inBangladeshrdquo Journal of the American Water Resources Associ-ation vol 51 no 3 pp 794ndash806 2015

[17] M R Rahman and H Lateh ldquoMeteorological drought inBangladesh assessing analysing and hazard mapping usingSPI GIS and monthly rainfall datardquo Environmental EarthSciences vol 75 no 12 2016

[18] L Han Q Zhang P Ma J Jia and J Wang ldquoThe spatial dis-tribution characteristics of a comprehensive drought risk indexin southwestern China and underlying causesrdquoTheoretical andApplied Climatology vol 124 no 3 pp 517ndash528 2016

[19] C J Kim M J Park and J H Lee ldquoAnalysis of climate changeimpacts on the spatial and frequency patterns of drought usinga potential drought hazard mapping approachrdquo InternationalJournal of Climatology vol 34 no 1 pp 61ndash80 2014

[20] B S Kim I G Chang J H Sung and H J Han ldquoProjectionin future drought hazard of South Korea based on RCP climatechange scenario 85 using SPEIrdquo Advances in Meteorology vol2016 Article ID 4148710 23 pages 2016

[21] Q Zhang and J Zhang ldquoDrought hazard assessment in typicalcorn cultivated areas of China at present and potential climatechangerdquo Natural Hazards vol 81 no 2 pp 1323ndash1331 2016

[22] L Hao X Zhang and S Liu ldquoRisk assessment to Chinarsquosagricultural drought disaster in county unitrdquo Natural Hazardsvol 61 no 2 pp 785ndash801 2012

[23] J Zhang ldquoRisk assessment of drought disaster in the maize-growing region of Songliao Plain Chinardquo Agriculture Ecosys-tems and Environment vol 102 no 2 pp 133ndash153 2004

12 Advances in Meteorology

[24] Y-N Hu Y-J Liu H-J Tang Y-L Xu and J Pan ldquoContribu-tion of drought to potential crop yield reduction in a wheat-maize rotation region in the North China plainrdquo Journal ofIntegrative Agriculture vol 13 no 7 pp 1509ndash1519 2014

[25] Z Chen and G Yang ldquoAnalysis of drought hazards in NorthChina distribution and interpretationrdquo Natural Hazards vol65 no 1 pp 279ndash294 2013

[26] H Kim J Park J Yoo and T-W Kim ldquoAssessment of droughthazard vulnerability and risk a case study for administra-tive districts in South Koreardquo Journal of Hydro-EnvironmentResearch vol 9 no 1 pp 28ndash35 2015

[27] A K Mishra V P Singh and V R Desai ldquoDrought charac-terization a probabilistic approachrdquo Stochastic EnvironmentalResearch and Risk Assessment vol 23 no 1 pp 41ndash55 2009

[28] J Sheffield and E F Wood ldquoProjected changes in droughtoccurrence under future global warming from multi-modelmulti-scenario IPCCAR4 simulationsrdquoClimate Dynamics vol31 no 1 pp 79ndash105 2008

[29] J Sheffield K M Andreadis E F Wood and D P LettenmaierldquoGlobal and continental drought in the second half of thetwentieth century severity-area-duration analysis and tempo-ral variability of large-scale eventsrdquo Journal of Climate vol 22no 8 pp 1962ndash1981 2009

[30] A Dai ldquoIncreasing drought under global warming in observa-tions and modelsrdquo Nature Climate Change vol 3 no 1 pp 52ndash58 2013

[31] E J Burke and S J Brown ldquoEvaluating uncertainties in theprojection of future droughtrdquo Journal of Hydrometeorology vol9 no 2 pp 292ndash299 2008

[32] H P Chen JQ Sun andX L Chen ldquoFuture changes of droughtand flood events in China under a global warming scenariordquoAtmospheric and Oceanic Science Letters vol 6 no 1 pp 8ndash132013

[33] LWangW Chen andW Zhou ldquoAssessment of future droughtin Southwest China based on CMIP5 multimodel projectionsrdquoAdvances in Atmospheric Sciences vol 31 no 5 pp 1035ndash10502014

[34] R Selvaraju and S Baas Climate Variability and ChangeAdaptation to Drought in Bangladesh A Resource Book andTraining Guide vol 9 Food and Agriculture OrganizationRome Italy 2007

[35] S M Abdullah and M Rahman ldquoInitiating rain water har-vest technology for climate change induced drought resilientagriculture scopes and challenges in Bangladeshrdquo Journal ofAgriculture and Environment for International Development vol109 no 2 pp 189ndash208 2015

[36] MoA Bangladesh Agriculture at a Glance Ministry of Agricul-ture Government of the Peoplersquos Republic of Bangladesh 2013httpwwwmoagovbdstatisticsbaghtm

[37] M R KarimM Ikeda andM Ishikawa ldquoRecent climate effectson seasonal rice yields in Bangladesh a statistical overviewrdquoJournal of Agricultural Science and Technology B vol 1 pp 950ndash963 2011

[38] U Habiba R Shaw and A W Hassan ldquoDrought Risk andReduction Approaches in Bangladeshrdquo in Disaster Risk Reduc-tionApproaches in Bangladesh Disaster RiskReduction pp 131ndash164 Springer Japan Tokyo Japan 2013

[39] S Shahid X Chen and M K Hazarika ldquoAssessment aridity ofBangladesh using geographic information systemrdquo GIS Devel-opment vol 9 no 12 pp 40ndash43 2005

[40] BanglapediaNational Encyclopedia of Bangladesh Asiatic Soci-ety of Bangladesh Dhaka Bangladesh 2003

[41] IPCC Climate Change 2014 IPCC Fifth Assessment Synthe-sis Report-Summary for Policymakers-an Assessment of Inter-Governmental Panel on Climate Change Cambridge UniversityPress Cambridge UK 2014

[42] M A Semenov and E M Barrow LARS-WG A StochasticWeather Generator for Use in Climate Impact Studies Version30 User Manual 2002

[43] M A Semenov R J Brooks E M Barrow and C W Richard-son ldquoComparison of the WGEN and LARS-WG stochasticweather generators for diverse climatesrdquo Climate Research vol10 no 2 pp 95ndash107 1998

[44] E Pasho J J Camarero M de Luis and S M Vicente-Serrano ldquoImpacts of drought at different time scales on forestgrowth across a wide climatic gradient in north-eastern SpainrdquoAgricultural and Forest Meteorology vol 151 no 12 pp 1800ndash1811 2011

[45] S M Vicente-Serrano S Beguerıa J Lorenzo-Lacruz et alldquoPerformance of drought indices for ecological agriculturaland hydrological applicationsrdquo Earth Interactions vol 16 no 102012

[46] V Potopova P Stepanek M Mozny L Turkott and J SoukupldquoPerformance of the standardised precipitation evapotran-spiration index at various lags for agricultural drought riskassessment in the Czech Republicrdquo Agricultural and ForestMeteorology vol 202 pp 26ndash38 2015

[47] C W Thornthwaite ldquoAn approach toward a rational classifica-tion of climaterdquo Geographical Review vol 38 no 1 pp 55ndash941948

[48] M A Abedin U Habiba and R Shaw ldquoImpacts of salinityarsenic and drought in south-western Bangladeshrdquo in Environ-ment Disaster Linkages R Shaw andT Phong Eds pp 165ndash193Emerald Publishers Bingley UK 2012

[49] A R M T Islam A Tasnuva S C Sarker M M RahmanM S H Mondal and M M U Islam ldquoDrought in NorthernBangladesh social agroecological impact and local perceptionrdquoInternational Journal of Ecosystem vol 4 no 3 pp 150ndash1582014

[50] M N Hossain S Chowdhury and S K Paul ldquoFarmer-level adaptation to climate change and agricultural droughtempirical evidences from the Barind region of BangladeshrdquoNatural Hazards vol 83 no 2 pp 1007ndash1026 2016

Submit your manuscripts athttpswwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ClimatologyJournal of

EcologyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EarthquakesJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom

Applied ampEnvironmentalSoil Science

Volume 2014

Mining

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal of

Geophysics

OceanographyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of Computational Environmental SciencesHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

GeochemistryHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Atmospheric SciencesInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

OceanographyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MineralogyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MeteorologyAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Paleontology JournalHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Geological ResearchJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Geology Advances in