Ecological Engineering 75 (2015) 303–307

Short communication

Drought dynamics and impacts on vegetation in China from 1982 to2011

Honglin Wang a, Aifang Chen a, Qianfeng Wang b, Bin He a,*aAcademy of College of Global Change and Earth System, Beijing Normal University, Beijing 100875, ChinabAcademy of Disaster Reduction and Emergency Management, Ministry of Civil Affairs and Ministry of Education, Beijing Normal University, Beijing 100875,China

A R T I C L E I N F O

Article history:Received 3 July 2014Received in revised form 20 November 2014Accepted 29 November 2014Available online xxx

Keywords:Climate changeDroughtSPIVegetationNDVIGrowing seasonChina

A B S T R A C T

We investigated the drought dynamics and their impacts on vegetation change in China from 1982 to2011 using the standard precipitation index (SPI) and the normalized difference vegetation index (NDVI)anomaly, which were calculated from meteorological and satellite-derived NDVI data, respectively. Thetrends in the change of SPI and vegetation were explored based on the non-parametric Mann–Kendall(MK) test and Sen’s slope test, and the relationship between these trends was examined. The results wereas follows: (1) For China as a whole, although the long term trend of drought-impacted areas changedlittle (�0.045% /10 a) over the past 30 years, Dry trends were identified in northeastern and southwesternChina. (2) The annual vegetation growth at the national scale showed an increasing trend, with a rate of0.008%/10 a from 1982 to 2011; cropland vegetation presented the largest increase in NDVI (p < 0.05). (3)Droughts that occurred during the growing season and pre-growing season both had large negativeimpacts on vegetation growth, and significant influences were found in northern China, especially in thenorthwestern area. Compared to the northern areas, the NDVI in southern China appeared to benefit fromwarming temperatures.

ã 2014 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Ecological Engineering

journal homepage: www.else vie r .com/ locat e/e coleng

1. Introduction

Many parts of the world have suffered from serious droughts inrecent years, which have greatly impacted socio–economicsystems and the environment (IPCC, 2001). Major drought eventshave raised the question of whether droughts are becoming morefrequent and more severe, as projected by climate change studies(Zhao and Running, 2010). Another concern is whether drought,which is one of the major disturbances to vegetation growth, mayweaken the ability of vegetation to function as a carbon sink (Xuet al., 2011). For the improved understanding of the vulnerability ofthe terrestrial carbon cycle and the sustainable use of naturalvegetation resources (Xu et al., 2012), it is of particular importanceto understand and assess the consequences of drought onvegetation at different scales.

The impacts of drought on vegetation have been extensivelystudied on regional and global scales. For instance, a recent studyindicated that droughts are counteracting the increase in global netprimary productivity (NPP) caused by global warming (Zhao and

* Corresponding author. Tel.: +86 15 810 327 627.E-mail address: hebin@bnu.edu.cn (B. He).

http://dx.doi.org/10.1016/j.ecoleng.2014.11.0630925-8574/ã 2014 Elsevier B.V. All rights reserved.

Running, 2010). On a regional scale, the severe 2003 drought inEurope caused great impacts on a variety of land-cover types, andmost ecosystems recovered to a normal state by early 2004(Gobron et al., 2005). Rainforests in the Amazon region play animportant role in the global carbon cycle, and extreme droughts inthis region in 2010 seemed to cause a significant increase in treemortality and carbon losses (Xu et al., 2011). Thus, serious droughtsare causing adverse effects on terrestrial ecosystems.

Vegetation growth has significantly increased over the lastthree decades in China (He et al., 2007). This can be mainlyattributed to national warming and ecological conservationprojects, such as the Three-North Shelter Forest Program (TNSFP)(Duan et al., 2011), the Beijing–Tianjin Sand Source ControlProgram (Wu et al., 2013), the Grain for Green Project (Zhang et al.,2012a), and small-scale regional ecological engineering (Huanget al., 2012). During the 20th century, China experienced a series ofdrought events (Xiao et al., 2009), and the increasing droughtstress associated with warming and reduced rainfall was found tocontribute to the decrease in the growing season’s normalizeddifference vegetation index (NDVI) in northern China after the1990s (Peng et al., 2011). In 2010, southwest China suffered asevere and sustained spring drought, which reduced the regionalannual gross primary productivity (GPP) and NPP by 65 and 45 Tg

304 H. Wang et al. / Ecological Engineering 75 (2015) 303–307

C/a, respectively, and both the annual GPP and NPP in 2010 werethe lowest during the period of 2000–2010 (Zhang et al., 2012b).Obviously, drought has become one of the most importantdisturbances to vegetation growth in China. Furthermore, droughtslower the water levels of rivers, reservoirs, and lakes, which limitswater availability for ecosystem and agricultural production andleads to ecological degradation and reduced food productivity(Piao et al., 2010; Zhang et al., 2012b; Yu et al., 2013).

Although previous studies have provided much informationon the relationship between drought and vegetation growth,many efforts mainly focused on the impacts of single droughtevents were concentrated on a regional scale. In addition, theresponses of vegetation to drought during the growing andnon-growing seasons are still not clear. Our main goal is to explorethe drought dynamics in China and their impacts on vegetationfrom 1982 to 2011. Firstly, the spatial-temporal characteristicsof drought are analyzed based on drought area and the trend ofdrought index. Secondly, the relationships between droughtvariations and vegetation growth are used to explore howa drought will impact vegetation growth. Our results willprovide scientific information for drought prediction and forestmanagement in China.

2. Datasets and methods

2.1. Datasets

The monthly precipitation (mm) data from 1982–2011 at752 meteorological stations in China were collected from the ChinaMeteorological Data Sharing Service. To avoid the possible effectsof artificial shifts in the data caused by the relocationsof measurement sites and equipment observation error, themeteorological data were checked for homogeneity with referenceto a previously used method (Ren et al., 2008). Stations with lessthan 30 years of data were also rejected (Yu et al., 2013). A total of603 stations were selected for this study according to the dataavailability criteria in China.

The NDVI is usually used for measuring vegetation growth. Inthis study, we used the National Oceanic and AtmosphericAdministration/Advanced Very High Resolution Radiometer(NOAA/AVHRR) NDVI dataset, which was produced by GlobalInventory Modeling and Mapping Studies (GIMMS) (Tucker et al.,2005). The dataset spans the period from 1982–2011 and has a0.083� spatial resolution and a 15-day interval. The GIMMS NDVIhas been corrected to remove non-vegetation effects, includingsensor degradation, inter-satellite differences, and volcanic aerosoleffects (Zhou et al., 2001). The monthly NDVI was derived from twoimages from each month using the maximum value composite(MVC) method (Holben 1986). Our analysis was mainly confined tothe growing season, which was defined as April–October and wasconstantly applied across the whole country (Zhou et al., 2001).Based on vegetation maps with a scale of 1:1,000,000 from theAtlas of China's Vegetation, the vegetation was classified intofour types: forest, shrub, grassland, and cropland (Zhao et al.,2011).

2.2. Methods

The standardized precipitation index (SPI) (McKee et al., 1993)was selected to assess drought variations in China. The SPI is awidely used drought index, which is derived from precipitationdata alone to determine a water deficit and surplus and can becalculated for short or long time scales (Paulo et al., 2003). Themulti-time scale SPI allowed us to analyze the relationship

between drought and vegetation at various time scales. Thecriteria for drought classification was referenced in McKee's study(McKee et al., 1993).

The non-parametric Mann–Kendall (MK) test was appliedfor the SPI trend tests. The MK test is a rank-based procedure,which is suitable for detecting non-linear trends (Kendall, 1975).It is frequently used for detecting trends in hydrologicaland meteorological time series (Hamed, 2008).Confidence probabilities of 95% (p < 0.05) was considered to besignificant.

The Sen slope estimator was applied to obtain the trend invegetation growth from 1982 to 2011 (Sen, 1968). This estimationdid not require the data to be distributed normally and has beenwidely applied in vegetation growth studies (Fernandes andLeblanc, 2005).

The bilinear interpolation method was used to extract the NDVIvalues for the station data based on grid vegetation data with aspatial resolution of 0.083� (Vu et al., 2012). The monthly NDVIsequence data were generated at station in China and spans theperiod from 1982 to 2011. Finally, the Pearson correlation analysiswas used to explore the relationship between the NDVI anomalyand drought (SPI less than �1.0) (Madden and Williams, 1978;Trenberth and Shea, 2005). The data processing and calculationswere conducted with the Interactive Data Language (IDL) programand the ArcGIS9.3 software.

3. Results

3.1. Temporal and spatial characteristics of drought

The inter-annual variation in the percentage of drought-impacted areas (SPI of less than �1.0) in China from 1982–2011 are shown in Fig. 1(a). Relatively large dry areas occurred inthe middle of the 1980s, the late 1990s, the early 2000s, and themost recent 1–2 years. The largest drought-impacted areapercentage (27.69%) occurred in 1986 and represented approxi-mately one-quarter of China's total territory. This value wasfollowed by 26.07% in 2001, 25.1% in 1997, and 22.51% in 2011.Despite the large inter-annual variability, the long-term trends inthe drought-impacted areas in China as a whole decreased slightly(by 0.045%/10 a) over the past 30 years.

The annual trends of SPI for each station in China were alsoinvestigated based on the MK test. The positive and negativeresults indicated trends towards wetter and drier conditions,respectively (Fig. 1(b)). The SPI trends showed great spatialvariability. Dry trends were identified in northeastern andsouthwestern China, but significant wet trends were detected inwestern China.

3.2. Trends in vegetation growth over the past 3 decades

Inter-annual variations in the NDVI for different vegetationtypes are shown in Fig. 2(a). The NDVI increased remarkably for allvegetation types (p < 0.05). The largest increases appeared incroplands and shrubs, with values of 0.015/10 a and 0.01/10 arespectively, followed by 0.007/10 a for grasslands and 0.006/10 afor forests. At the national scale, the average NVDI for all annualvegetation in China demonstrated a significant increasing trend,with a rate of 0.007/10 a. The annual NDVI trend patterns weredramatically spatially heterogeneous (Fig. 2(b)). Significantincreasing trends in vegetation growth occurred in central, eastern,and southern China, and significant decreasing trends occurred inthe northwestern and northeastern areas and some parts of thesouthwestern area.

Fig. 1. Percentage of drought-impacted areas (a) and annual SPI trends from 1982 to 2011 (b).

H. Wang et al. / Ecological Engineering 75 (2015) 303–307 305

3.3. Relationship between drought and vegetation

To assess the impact of droughts occurring in the growingseason (April–October) and pre-growing season (November of theprevious year to March) on vegetation, we calculated therelationships between the growing season SPI and NDVI anomaly,and between the pre-growing season SPI and growing season NDVIanomaly respectively, as shown in Fig. 3. For both conditions,significant positive relationships mainly occurred in the north-western and northern areas of China, which are arid and semi-aridregions. This indicated that droughts occurring during the growingseason or pre-growing season should have large adverse impactson vegetation growth. In addition, more stations had significantpositive relationships between growing season SPI and NDVIcompared with the pre-growing season (60 vs. 48). This indicatedthat droughts occurring in the growing season might have moreserious influences on vegetation growth than those occurring inthe pre-growing season. It is worthy to note that the vegetation innortheastern China is very sensitive to pre-growing seasondrought, but it is less sensitive to growing season drought.

4. Discussion

The trends in the drought-impacted areas showed slightchanges and were essentially consistent with the results calculatedusing Palmer Drought Severity Index (PDSI), which presented an

Fig. 2. Inter-annual variations of annual NDVI (a) and the spa

increasing trend of 0.50%/10 a from 1951 to 2003 (Zou, 2005).However, a significant upward trend with a rate of 3.7%/10 a from1951 to 2010 (Yu et al., 2013) was revealed based on another widelyused drought index, the Standardized Precipitation Evapotranspi-ration Index (SPEI). Through investigating the trend in the SPI, wefound that dry trends occurred in the northeast and southwestregions of China, which were similar to the areas that experiencedprecipitation decreases (Yu et al., 2013). The SPI values werecalculated solely based on precipitation data; thus, the environ-mental factors that impact the distribution of precipitation,including latitude, longitude, and topography, may also have aninfluence on the spatial patterns of the SPI (Zhai et al., 2005; Zhangand Feng, 2010).

A significant increasing trend in the annual vegetation growthin China was found in our study. This result was similar to that ofPeng's study, in which the growing season (April–October) NDVIwas found to significantly increase by 0.0007 year�1 from 1982 to2010 on a national scale (Peng et al., 2011). The spatial patterns ofvegetation growth changes were generally consistent with otherfindings on vegetation changes in China (He et al., 2007; Xu et al.,2012). In northern China, the vegetation growth displayed adecreasing trend, which was attributed to warming temperaturesand the related increasing evapotranspiration rates and soil waterdeficits (Duan et al., 2011). The relationship between the growingseason NDVI and temperature anomalies is shown in Fig. 4. Thehigh negative correlations mainly in the north indicated that

tial distribution of annual NDVI trends (b) for 1982–2011.

Fig. 3. The relationship between growing season (April–October) SPI and NDVI (a) and between the pre-growing season SPI (November of the previous year to March) andgrowing season NDVI anomaly (b).

306 H. Wang et al. / Ecological Engineering 75 (2015) 303–307

increasing temperatures could lead to adverse effects on vegeta-tion growth. However, increasing trends in NDVI were also found insmall parts of North China, such as the western part of InnerMongolia. The possible reasons for this observation are thefollowing: (1) the negative effects of increasing temperature onvegetation growth were offset by increasing precipitation; (2) theencroachment of trees and shrubs into grasslands that are affectedby climate, soil formations, and other factors promoted vegetationgrowth, and thus affected the aboveground net primary produc-tivity (ANPP) and ecosystem carbon (C) and nitrogen (N) pools(Huang et al., 2012); (3) ecological engineering benefited vegeta-tion growth in this region (Peng et al., 2011). Increasing trends invegetation were observed in southern China. A widely acceptedviewpoint is that vegetation in this region is mainly controlled bytemperature. Warming temperatures have prolonged the growingseason, improved photosynthetic efficiency, and increased vege-tation productivity (Duan et al., 2011). As shown in Fig. 4, higherpositive correlations were mainly found in the south.

Significant positive correlations between the SPI and NDVIanomaly were identified for both the growing and pre-growingseasons in north China (Wang et al., 2010), implying thatvegetation growth was affected not only by the growing seasonprecipitation deficit, but also by the pre-growing season droughtconditions. These results are not difficult to understand, as manyprevious studies assert that precipitation is the main factor

Fig. 4. Correlation between growing season NDVI and temperature anomalies.

limiting vegetation growth (Zhao et al., 2011). Although thevegetation could be restored to a normal status after the end of adrought (Zhang et al., 2012b), but long duration and frequentdroughts will inevitably lead to the vegetation degradation.Significant negative correlations between the SPI and NDVIanomaly occurred in the southern region, indicating that theprecipitation deficits favored vegetation growth. This was mainlydue to the resilience and restorability of vegetation to droughtstress. Previous studies have confirmed that resilience of vegeta-tion to drought stress is stronger in humid regions than in otherregions (Vicente-Serrano et al., 2013). In addition, the cloudcoverage in southern China is markedly larger than that innorthern China (less cloud coverage for more solar radiation) (Heet al., 2007). During drought periods the increased radiation due todecreasing cloud coverage may promote vegetation growth(Saleska et al., 2007).

Overall, drought is an important disturbance to vegetationgrowth in China, especially for North China. Other factors,including the influence of climate change, ecological engineering,the resilience and restorability of vegetation, the encroachment ofother biome types, etc., contribute to the complex interactionprocesses of vegetation growth and drought.

5. Conclusion

We conclude that the variation trends in drought-impactedareas were slight (�0.045%/10 a) during 1982–2011; droughtsoccurring during the growing season and pre-growing season bothhad great impacts on the NDVI; and a significant influence ofdrought on vegetation was found in northern China, especially inthe northwest. Our study on drought dynamics in China can guideagricultural production for the continued development of adapta-tion strategies to protect vulnerable ecosystems and to ensureagricultural security. The findings on the relationship betweendrought and vegetation growth can also provide a vital scientificbasis for China's drought predication and forest management.

Acknowledgments

This work was financially supported by the National BasicResearch Development Program of China (grant no.2012CB95570001 and 2011CB952001) and the NationalNatural Science Foundation of China (grant no. 41301076) andthe State Key Program of National Natural Science of China(grant no. 41330527).

H. Wang et al. / Ecological Engineering 75 (2015) 303–307 307

References

Duan, H., Yan, C., Tsunekawa, A., Song, X., Li, S., Xie, J., 2011. Assessing vegetationdynamics in the Three-North Shelter Forest region of China using AVHRR NDVIdata. Environ. Earth Sci. 64, 1011–1020.

Fernandes, R., Leblanc, S.G., 2005. Parametric (modified least squares) and non-parametric (Theil–Sen) linear regressions for predicting biophysical parametersin the presence of measurement errors. Remote Sens. Environ. 95, 303–316.

Gobron, N., Pinty, B., Mélin, F., Taberner, M., Verstraete, M.M., Belward, A., Lavergne,T., Widlowski, J.L., 2005. The state of vegetation in Europe following the2003 drought. Int. J. Remote Sens. 26, 2013–2020.

Hamed, K.H., 2008. Trend detection in hydrologic data: the Mann–Kendall trend testunder the scaling hypothesis. J. Hydrol. 349, 350–363.

He, Y., Dong, W., Guo, X., Dan, L., 2007. Terrestrial growth in China and itsrelationship with climate based on the MODIS data. Acta Ecol. Sin. 27,5086–5092.

Holben, B.N., 1986. Characteristics of maximum-value composite images fromtemporal AVHRR data. Int. J. Remote Sens. 7, 1417–1434.

Huang, L., Shao, Q., Liu, J., 2012. Forest restoration to achieve both ecological andeconomic progress, Poyang Lake basin, China. Ecol. Eng. 44, 53–60.

IPCC, 2001. Climate Change 2001: Impacts, Adapation and Vulnerability of ClimateChange, Working Group II Report. Cambrige University Press, London.

Kendall, M., 1975. Rank Correlation Methods. Grif, London.Madden, R.A., Williams, J., 1978. The correlation between temperature and

precipitation in the United States and Europe. Mon. Weather Rev. 106, 142–147.McKee Thomas, B., Nolan Doesken, J., Kleist, J., 1993. The relationship of drought

frequency and duration to time scales. Proceedings of the 8th Conference onApplied Climatology Vol. 17. No. 22, Boston, MA: American MeteorologicalSociety,.

Paulo, A.A., Pereira, L.S., Matias, P.G., 2003. Analysis of local and regional droughts inSouthern Portugal using the theory of runs and the standardised precipitationindex. Pages 55–78. In: Rossi, G., Cancelliere, A., Pereira, L., Oweis, T., Shatanawi,M., Zairi, A. (Eds.), Tools for Drought Mitigation in Mediterranean Regions.Springer, Netherlands.

Peng, S., Chen, A., Xu, L., Cao, C., Fang, J., Myneni, R.B., Pinzon, J.E., Tucker, C.J., Piao, S.,2011. Recent change of vegetation growth trend in China. Environ. Res. Lett. 6,044027.

Piao, S., Ciais, P., Huang, Y., Shen, Z., Peng, S., Li, J., Zhou, L., Liu, H., Ma, Y., Ding, Y.,Friedlingstein, P., Liu, C., Tan, K., Yu, Y., Zhang, T., Fang, J., 2010. The impacts ofclimate change on water resources and agriculture in China. Nature 467,43–51.

Ren, G., Zhou, Y., Chu, Z., Zhou, J., Zhang, A., Guo, J., Liu, X., 2008. Urbanization effectson observed surface air temperature trends in North China. J. Clim. 21,1333–1348.

Saleska, S.R., Didan, K., Huete, A.R., da Rocha, H.R., 2007. Amazon forests green-upduring 2005 drought. Science 318, 612.

Sen, P.K., 1968. Estimates of the regression coefficient based on Kendall's Tau. J. Am.Stat. Assoc. 63, 1379–1389.

Trenberth, K.E., Shea, D.J., 2005. Relationships between precipitation and surfacetemperature. Environ. Res. Lett. 32, L14703.

Tucker, C.J., Pinzon, J.E., Brown, M.E., Slayback, D.A., Pak, E.W., Mahoney, R., Vermote,E.F., El Saleous, N., 2005. An extended AVHRR 8 km NDVI dataset compatiblewith MODIS and SPOT vegetation NDVI data. Int. J. Remote Sens. 26, 4485–4498.

Vicente-Serrano, S.M., Gouveia, C., Camarero, J.J., Beguería, S., Trigo, R., López-Moreno, J.I., Azorín-Molina, C., Pasho, E., Lorenzo-Lacruz, J., Revuelto, J., Morán-Tejeda, E., Sanchez-Lorenzo, A., 2013. Response of vegetation to drought time-scales across global land biomes. Proc. Nat. Acad. Sci. 110, 52–57.

Vu, M.T., Raghavan, S.V., Liong, S.Y., 2012. SWAT use of gridded observations forsimulating runoff–a Vietnam river basin study. Hydrol. Earth Syst. Sci. 16,2801–2811.

Wang, W., Wang, W.-j., Li, J.-s., Wu, H., Xu, C., Liu, T., 2010. The impact of sustaineddrought on vegetation ecosystem in Southwest China based on remote sensing.Procedia Environ. Sci. 2, 1679–1691.

Xiao, J., Zhuang, Q., Liang, E., Shao, X., McGuire, A.D., Moody, A., Kicklighter, D.W.,Melillo, J.M., 2009. Twentieth-century droughts and their impacts on terrestrialcarbon cycling in China. Earth Interact. 13, 1–31.

Xu, L., Samanta, A., Costa, M.H., Ganguly, S., Nemani, R.R., Myneni, R.B., 2011.Widespread decline in greenness of Amazonian vegetation due to the2010 drought. Geophys. Res. Lett. 38.

Xu, X., Piao, S., Wang, X., Chen, A., Ciais, P., Myneni, R.B., 2012. Spatio-temporalpatterns of the area experiencing negative vegetation growth anomalies inChina over the last three decades. Environ. Res. Lett. 7, 035701.

Yu, M., Li, Q., Hayes, M.J., Svoboda, M.D., Heim, R.R., 2013. Are droughts becomingmore frequent or severe in china based on the standardized precipitationevapotranspiration index: 1951–2010? Int. J. Climatol..

Zhai, P., Zhang, X., Wan, H., Pan, X., 2005. Trends in total precipitation and frequencyof daily precipitation extremes over China. J. Clim. 18, 1096–1108.

Zhang, H., Feng, L., 2010. Characteristic of spatio-temporal variation of precipitationin North China in recent 50 years. J. Nat. Resour. 270–279.

Zhang, G., Dong, J., Xiao, X., Hu, Z., Sheldon, S., 2012a. Effectiveness of ecologicalrestoration projects in Horqin Sandy Land: China based on SPOT-VGT NDVI data.Ecol. Eng. 38, 20–29.

Zhang, L., Xiao, J., Li, J., Wang, K., Lei, L., Guo, H., 2012b. The 2010 spring droughtreduced primary productivity in southwestern China. Environ. Res. Lett. 7,045706.

Zhao, M., Running, S.W., 2010. Drought-induced reduction in global terrestrial netprimary production from 2000 through 2009. Science 329, 940–943.

Zhao, X., Tan, K., Zhao, S., Fang, J., 2011. Changing climate affects vegetation growthin the arid region of the northwestern China. J. Arid Environ. 75, 946–952.

Zhou, L., Tucker, C.J., Kaufmann, R.K., Slayback, D., Shabanov, N.V., Myneni, R.B.,2001. Our goal is mainly to explore the relationship between drought. J.Geophys. Res. 106, 20069–20083.