2012 AGU Chapman Conference on Remote Sensing of the ...

seasonal melt pattern and thus to distinguish theoutstanding aerosol-induced snowmelt signal. Results fromthis observational work are expected to provide betterunderstanding of the radiative impact of aerosols over snowsurface, especially its role in the Himalayan hydroglacialogicalvariability.Huete, Alfredo R.Shifts in Rainfall Use Efficiency and PrimaryProduction along a Savanna Aridity GradientAssessed Using Satellite and Flux Tower Time SeriesDataHuete, Alfredo R. 1 ; Eamus, Derek 1 ; Ponce, Guillermo E. 3 ;Moran, M. S. 2 ; McVicar, Tim R. 4 ; Donohue, Randall 4 ;Restrepo-Coupe, Natalia 1 ; Davies, Kevin 1 ; Glenn, Edward P. 3 ;Cleverly, James 1 ; Boulain, Nicolas 1 ; Beringer, Jason 5 ; Hutley,Lindsay 61. Univ of Technology, Sydney, NSW, Australia2. USDA-ARS, Tucson, AZ, USA3. University of Arizona, Tucson, AZ, USA4. CSIRO Land and Water, Canberra, ACT, Australia5. Monash University, Clayton, VIC, Australia6. Charles Darwin University, Darwin, NT, AustraliaAustralia’s climate is extremely variable with interannualrainfall at any location varying up to eight-fold. In thenorthern tropical savannas there is also significantmonsoonal rainfall variability with pronounced seasonal dryperiods. Understanding water and productivity relationshipsrepresent key issues in climate change models that aim topredict how carbon and water relationships will shift withprojected changes in the frequency, timing, amount andintensity of rainfall. The goal of this study was to investigatethe relationships of above-ground net primary production(ANPP) with rainfall variability along the Northern AustraliaTropical Transect (NATT). We combined 11 years of MODISenhanced vegetation index (EVI) with rainfall data from theTropical Rainfall Monitoring Mission (TRMM) to assesslarge area spatial and temporal patterns in above-groundvegetation productivity (ANPP) and rainfall use efficiency(RUE), defined as ANPP divided by annual rainfall. ANPPvalues were retrieved by (1) coupling seasonal EVI values at16-day increments with tower flux measurements of grossprimary production (GPP) at 3 tower sites along the ariditygradient, followed by (2) annual integration of EVI values(iEVI) from baseline values with zero GPP fluxes, and (3)adjustments for ecosystem respiration. Rainfall useefficiencies were computed for each year as the slope of theiEVI by the annual rainfall. We found strong cross-siteconvergence of seasonal and interannual satellite EVI valueswith tower GPP fluxes at the three sites. Baseline EVI valuesenabled separation of tree and grass ANPP. As expected,positive curvilinear relationships were found between iEVIand annual rainfall with decreasing sensitivity of iEVI toadditional rainfall at the more humid regions of thetransect. We found ANPP values decreased from the humidnorthern savannas to the more southern arid portions,however, RUE increased along the same transect from north76to south resulting in the highest levels of ANPP per unitrainfall in the more arid savannas. Overall, the wet and drysavannas converged to a common and maximum RUE, orRUEmax, when plotted using the driest year at each pixel.However, the most humid northern tropical savannasyielded unexpectedly lower RUE and ANPP values, partlyattributed to a deficiency in tree leaf area to capture light forphotosynthesis. Rainfall and ANPP variability across thetransect were highest over the more grassland dominatedregions of the savanna, or lowest tree-grass ratiosdemonstrating a higher sensitivity of grassland biomes toclimate change relative to the more woody dominatedsavanna and forest biomes. The results of this study suggestthat commonly accepted patterns of ANPP response torainfall may not apply in tropical wet-dry savannas underpredicted climate change.Huffman, George J.Upgrades to the Real-Time TMPAHuffman, George J. 2, 1 ; Bolvin, David T. 2, 1 ; Nelkin, Eric J. 2, 1 ;Adler, Robert F. 3 ; Stocker, Erich F. 11. ESD, NASA GSFC, Greenbelt, MD, USA2. Science Systems and Applications, Inc., Lanham, MD,USA3. ESSIC, Univ. of Maryland, College Park, College Park,MD, USAThe TRMM Multi-satellite Precipitation Analysis(TMPA) provides 0.25°x0.25° 3-hourly estimates ofprecipitation in the latitude band 50°N-50°S in two productsets. First, it is computed 6-9 hours after real time usingprecipitation estimates from imager and sounder passivemicrowavesatellite instruments, and geosynchronous-orbitIR (geo-IR) data, all intercalibrated to a single TRMM-basedstandard, the TMI-GPROF product. Second, the TMPA iscomputed about two months after the end of each calendarmonth (in the current Version 7) using the same satellitedata as in the real time, but calibrated to the TRMMCombined Instrument (TCI) product, with input frommonthly rain gauge analyses. Respectively, these two versionsare referred to as the real-time TMPA-RT (3B42RT) andresearch TMPA (3B42) products. Recently, these productshave been revised to Version 7 (after a lengthy delay due toinput data issues). One key change is that we have developeda computationally feasible scheme for reprocessing theTMPA-RT for the entire TRMM record. Users stronglyadvocated this innovation, and early results will be shown.The reprocessing is important not only for providing alonger record, but also for incorporating more-consistentarchives of input data. The time series of the RT andresearch products are similar, although the research productperforms better both in terms of bias and random error.Over land, this improvement is due to both the TCIcalibration and the use of gauges, while over ocean it is theresult of TCI calibration alone. Climatological calibration forthese factors was instituted in the RT processing duringVersion 6 and it has been continued in Version 7, with thenecessary recalibrations to the Version 7 production data.The second part of the study will summarize early results on