improving modis thermal emissive bands calibration · improving modis thermal emissive bands...

Improving MODIS Thermal Emissive Bands CalibrationJack Xiong1, G. Keller2 and T. Wilson21Biospheric Sciences, NASA GSFC , 2SSAI

Figure 1 Figure 2

Earth Sciences Division – Hydrosphere, Biosphere, and Geophysics

EOS

Improving MODIS thermal emissive bands on-orbit calibration and image quality using crosstalkcorrection algorithms and correction coefficients derived from on-orbit lunar observations.


EOS

Name: Jack Xiong, Biospheric Sciences, NASA GSFCE-mail: [email protected]: 301-614-5957

References:Xiong, X., Aisheng Wu, B. N. Wenny, et al. 2015. "Terra and Aqua MODIS Thermal Emissive Bands On-Orbit Calibration and Performance." IEEE Transactions on Geoscience and Remote Sensing, 53 (10): 5709-5721 [10.1109/tgrs.2015.2428198]Wilson, T.., A. Wu, A.. Shrestha, X. Geng, Z. Wang, C. Moeller, R.. Frey, and X. Xiong, 2017. "Development and Implementation of an Electronic Crosstalk Correction for Bands 27–30 in Terra MODIS Collection 6." Remote Sensing, vol. 9(6), issue 569 [10.3390/rs9060569]. Keller, G.., Z. Wang, A. Wu, and X. Xiong, 2017. "Aqua MODIS Band 24 Crosstalk Striping." IEEE Geoscience and Remote Sensing Letters, vol. 14, issue 4, pp. 475-479 [10.1109/LGRS.2016.2647441]

Data Sources: All sensor calibration raw data and the data used to generate the images are from NASA GSFC Level 1 and Atmosphere Archive and Distribution System (LAADS). The L1B calibration and crosstalk correction algorithms and their corresponding coefficients are derived by the NASA MODIS Characterization Support Team (MCST).

Technical Description of Figures:Figure 1: Sample L1B images from Aqua MODIS band 24 (data granules from 2016). Top: before applying crosstalk correction, showing striping; bottom: after crosstalk correction with striping significantly reduced.Figure 2: Intensity profiles corresponding, from top to bottom, to the images in Figure 1, from left to right (black/red is before/after crosstalk correction). The profiles were extracted from the regions marked with red vertical lines in Figure 1.

Scientific significance, societal relevance, and relationships to future missions: Terra and Aqua MODIS have successfully operated for more than 17 and 15 years since their launch in December 1999 and May 2002, respectively. MODIS data products have been widely used for studies of many critical environmental parameters of the earth’s system. Electronic crosstalk was initially identified pre-launch in MODIS thermal emissive bands (TEB) and, over the course of the mission, its impact have become more server or non-negligible for several bands. Improved crosstalk correction algorithms have been developed recently and extensively tested. The implementation of crosstalk correction will be made in Terra MODIS C6.1 as an effort to improve its LWIR spectral bands (27-30) calibration and image quality. Methodologies developed and lessons from MODIS crosstalk characterization have potential applications for other earth-observing sensors, such as JPSS VIIRS and GOES-R ABI.

AMSR2 Soil Moisture Development and ValidationRajat Bindlish1, Tom Jackson2, Michael Cosh2 and Steven Chan3

1Hydrological Sciences, NASA GSFC, 2USDA ARS, 3NASA JPL

The AMSR2 mission provides long-term continuity toAMSR-E observations and an opportunity to develop aclimate data record. The AMSR2 Single Channel (X-band) Algorithm (SCA) soil moisture product performswell over low to moderately vegetated areas andoutperforms the JAXA and LPRM products in terms ofbias and ubRMSE.

The advantages of using lower frequency (L-band)channels for soil moisture retrieval are reflected insignificantly better performance metrics for SMAP thanAMSR2 at the same core validation sites.

Table 1


AMSR2 Soil Moisture (SCA) (August 2015)

Core SiteAccuracy Metrics

JAXA SCA LPRMubRMSE(m3/m3)

Bias (m3/m3)

RMSE (m3/m3) R

ubRMSE(m3/m3)

Bias (m3/m3)

RMSE (m3/m3) R

ubRMSE(m3/m3)

Bias (m3/m3)

RMSE (m3/m3) R

AMSR2 Des. (1:30 AM) 0.059 -0.089 0.111 0.502 0.055 -0.047 0.080 0.569 0.088 0.100 0.137 0.601

AMSR2 Asc. (1:30 PM) 0.057 -0.081 0.102 0.541 0.056 -0.046 0.081 0.586 0.090 0.045 0.104 0.540

SMAP Level 2 Passive ProductSMAP Des.(6:00 AM) - - - - 0.037 -0.014 0.052 0.822 - - - -

SMAP Asc. (6:00 PM) - - - - 0.039 -0.028 0.061 0.795 - - - -

Figure 1

Name: Rajat Bindlish, Hydrological Sciences, NASA GSFC E-mail: [email protected]: 301-286-8753

References:R. Bindlish, M. H. Cosh, T. J. Jackson, T. Koike, H. Fujii, S. Chan, J. Asanuma, A. Berg, D. D. Bosch, T. Caldwell, C. Holifield Collins, H. McNairn,

J. Martínez-Fernández, J. Prueger, T. Rowlandson, M. Seyfried, P. Starks, M. Thibeault, R. Van Der Velde, J. P. Walker, E. Coopersmith. GCOM-W AMSR2 Soil Moisture Product Validation Using Core Validation Sites. (under review).

Chan, S., R. Bindlish, P. O’Neill, E. Njoku, T. Jackson, A. Colliander, F. Chen, M. Bürgin, S. Dunbar, J. Piepmeier, S. Yueh, D. Entekhabi, M. Cosh, T. Caldwell, J. Walker, X. Wu, A. Berg, T. Rowlandson, A. Pacheco, H. McNairn, M. Thibeault, J. Martínez-Fernández, Á. González-Zamora, M. Seyfried, D. Bosch, P. Starks, D. Goodrich, J. Prueger, M. Palecki, E. Small, J.C. Calvet, W. Crow, and Y. Kerr, “Assessment of the SMAP Level 2 passive soil moisture product,” IEEE Trans. Geosci. Remote Sens., 54: 4994-5007, 2016.

Data Sources: The Advanced Microwave Scanning Radiometer 2 (AMSR2) is part of JAXA’s Global Change Observation Mission-Water (GCOM-W) mission. GCOM-W is part of the A-train constellation with an sun-synchronous 1:30 AM/PM orbit. Passive microwave (X-band) imagery from JAXA AMSR2 are used for the SCA product (Figure 1). Data from JAXA AMSR2 and NASA SMAP are used in the in situ assessment of the different soil moisture products. In situ soil moisture data is provided by various US and international partners (15 CVS sites).

Technical Description: Figure 1: AMSR2 soil moisture estimates using the Single Channel Algorithm (SCA) for August 2015 (descending orbits-1:30 AM). Areas with dense vegetation and snow/frozen are masked. Arid areas (western US, Sahara, Southern Africa, Middle East and Central Australia) have low soil moisture. Areas with higher vegetation have greater soil vegetation have higher soil moisture. The onset of monsoon over Indian sub-continent and South-east Asia can be seen in the figure. Table 1: AMSR2 Ascending (1:30 AM and 1:30 PM) performance statistics for the three soil moisture products, Japanese Space Agency (JAXA), Single Channel Algorithm (SCA), and Land Parameter Retrieval Model (LPRM). Performance of the standard SMAP passive only (L2_SM_P) is also shown in the table.

Scientific significance, societal relevance, and relationships to future missions: Soil moisture plays a critical role in linking Earth’s water, energy and carbon cycles, and is critical to a large number of applications with high societal benefits. AMSR2 is a follow-on instrument to AMSR-E and provides an opportunity to develop a long-term soil moisture data using X-band microwave observations. The SCA performs well over low and moderately vegetated areas. JAXA and SCA products had a similar ubRMSE that met the target accuracy requirements for AMSR2. The assessment results from ascending and descending orbits are similar and both observations can be used in hydrologic applications. The SMAP passive only (L2SM_P) results using lower frequency (L-band) are significantly better than AMSR2 at the same core validation sites. The use of L-band observations provides an ability to estimate soil moisture over areas with higher vegetation and is the optimum frequency for estimating surface soil moisture. L-band observations are available since 2010 only and the future of L-band is unknown beyond SMOS and SMAP. Thus, X-band observations are our only current opportunity to estimate a long-term soil moisture climate data record.


Figure 1The SeaBASS bio-optical archive within the OceanEcology Lab has released, as a part of the award-winning Ocean Color software suite: SeaDAS (v7.4),satellite validation data search and match-up tools.

These software tools take in situ bio-opticalmeasurements, with requisite time/locationmetadata, and find coincident ocean color satelliteretrievals and create satellite extracts matched up toin situ measurements, after applying necessary QCcriteria to the data (Bailey and Werdell, 2006). Thesetools provide researchers in the scientific communitya way to conduct their own satellite validationstudies and customize their data searches and QCfilters in a simple 2-step process.

Satellite Validation Search, Extract, and Match-up ToolsJoel Scott1,2, Chris Proctor3,2, Jason Lefler4,2

1SAIC, 2 Ocean Ecology, NASA GSFC, 3SSAI, 4JHT, Inc.

A

A

B

B

C

C


Name: Joel Scott, Ocean Ecology, NASA GSFC / SAICE-mail: [email protected]: 301-286-3147

References:

Bailey, S., P.J. Werdell, 2006, A multi-sensor approach for the on-orbit validation of ocean color satellite data products, Rem. Sens. Environ. , vol 102, pp 12-23.

Balch, B., et al, 2014, AMT24 cruise data set, doi: 10.5067/SeaBASS/AMT/DATA001, https://seabass.gsfc.nasa.gov/archive/BIGELOW/BALCH/AMT/archive/amt24-merged-SAS-flow-ac9.out

SeBASS bio-optical data archive: https://seabass.gsfc.nasa.gov/

SeaDAS software suite website: https://seadas.gsfc.nasa.gov/

Data Sources: Satellite derived ocean color imagery from MODIS-Aqua, MODIS-Terra, and Suomi-NPP VIIRS.In situ bio-optical measurements Barney Balch, 2014 (DOI: 10.5067/SeaBASS/AMT/DATA001) flow-through system on the Atlantic Meridional Transect 24 (AMT24) cruise.SeaDAS (v7.4) Ocean Color software suite, https://seadas.gsfc.nasa.gov/, last accessed on 2017-06-30.

Technical Description of Figures: Figure 1: The images show log-log 2-dimensional histograms, also known as joint-Probability Density Functions, depicting in situ fluorometricchlorophyll-a measurements collected on the AMT24 cruise via the ship’s flow-through system by Barney Balch’s group at the Bigelow Laboratory for Ocean Sciences, submitted to and archived at NASA GSFC’s SeaBASS bio-optical archive, versus satellite retrieved chlorophyll-a concentrations from MODIS-Aqua (A), MODIS-Terra (B), and Suomi-NPP VIIRS (C). These validation matchups were created from a single in situ SeaBASS-format data file containing the in situ measurements with time and location metadata via the satellite validation match-up tools that are provided as a part of SeaDAS v7.4 and higher. These match-up tools (fd_matchup and mk_matchup) search for coincident ocean color satellite Level-2 granules via NASA’s EarthData Search and Common Metadata Repository APIs, fetch the relevant satellite data granules, and create satellite data extracts to compare to the in situ measurments, applying the match-up methodology and QC criteria recommended by Bailey and Werdell (2006). Figure 2: The locations of the MODIS-Aqua (A), MODIS-Terra (B), and Suomi-NPP VIIRS (C) match-ups to the in situ data are also plotted for reference.

Scientific significance, societal relevance, and relationships to future missions: Comparing in situ measurements to satellite-retrieved bio-optical properties, and distributing tools for the scientific community to conduct their own studies and comparisons, assesses the accuracy of the satellite retrievals and derived products, providing information that can then be used to improve existing satellite retrieval algorithms or develop new ones. Thereby, advancing humanity’s understanding of and our interactions with the Earth’s oceans and marine resources, which are critical to supporting life on this planet.


Monitoring land surface albedo and vegetation dynamics Zhuosen Wang1, Miguel O. Román1, Crystal B. Schaaf2, and Jeffrey G. Masek3

1Code 619, NASA/GSFC, 2University of Massachusetts Boston , 3Code 618, NASA/GSFC

Earth Sciences Division – Terrestrial Information Systems Laboratory

Synthetic Enhanced Vegetation Index (EVI) at Harvard Forest

EVI

Day of Year (DOY) in 2007

The synthetic 30m daily time series vegetation indices (EVI) and albedo generated from 500 m MODIS Collection V006 daily BRDF/NBAR/albedo products and 30 m Landsat 5 data provide much greater spatial detail than the MODIS data, especially over more heterogeneous surfaces, which improves the efforts to characterize and monitor the spatial and temporal variation across species and communities.

Harvard Forest

(A) (B)

Name: Zhuosen Wang (619), Miguel O. Román (619), Crystal B. Schaaf (Umass Boston), and Jeffrey G. Masek (618) E-mail: [email protected], [email protected], [email protected], [email protected]: 301-614-5140, 301-614-5498, 508-654-5554, 301-614-6629

References:Wang, Z., Schaaf, C. B., Sun, Q., Kim, J., Erb, A. M., Gao, F., … Papuga, S. A. (2017). Monitoring land surface albedo and vegetation dynamics using high spatial and temporal resolution synthetic time series from Landsat and the MODIS BRDF/NBAR/albedo product. International Journal of Applied Earth Observation and Geoinformation, 59, 104–117. http://doi.org/10.1016/j.jag.2017.03.008

Gao, F., Masek, J., Schwaller, M., & Hall, F. (2006). On the blending of the Landsat and MODIS surface reflectance: predicting daily Landsat surface reflectance. IEEE Transactions on Geoscience and Remote Sensing, 44(8), 2207–2218. http://doi.org/10.1109/TGRS.2006.872081

Data Sources: Landsat surface reflectance were downloaded from USGS GloVis. Collection V006 MODIS BRDF/NBAR/albedo products were downloaded from NASA LAADS website. Ground Harvard forest phenology metrics were measured by John O’Keefe from Harvard University .

Technical Description of Figures:

Graph (A) : Time series synthetic (red “+”) and logistic model fit (blue line) Enhanced Vegetation Index (EVI) at Harvard forest flux tower in 2007. The green line shows the ground measured forest green-up date.

Graph (B): The time series synthetic and MODIS shortwave (SW) broadband blue sky albedo at Harvard Forest subset (16 km by 14 km) on DOY 95, 125, 135, 160 and 305 in2007..

Scientific significance, societal relevance, and relationships to future missions: The MODIS-Landsat coupled synthetic 30m daily albedo and EVI times series are able to capture the land surface dynamics at high spatial resolution. Such a capability lays the ground work for long-term monitoring of vegetation phenology at the stand scale in response to environment change, disturbance regimes, and other drivers. The heterogeneity analyses in this study indicate that the range of EVI and albedo within moderate spatial resolution grids is very large and higher spatial resolution vegetation index and albedo values are necessary to understand how individual vegetation types are responding to environmental forcing. The observations of higher resolution surface phenology and energy change from the synthetic time series data will be evaluated with fine spatial resolution albedo from NASA Multi AngLe Imaging Bidirectional Reflectance Distribution Function sUAS (MALIBU) observations and continued with the newer generation of satellites including Landsat 8 Operational Land Imager (OLI), Sentinel-2A/B Multi Spectral Instrument (MSI), and Suomi-NPP Visible Infrared Imager Radiometer Suite (VIIRS) satellite sensors.

Earth Sciences Division – Terrestrial Information Systems Laboratory

improving modis thermal emissive bands calibration · improving modis thermal emissive bands...

Documents