Sensitivity of Three Vegetation Indices to Cimate and Soil moisture at a Tallgrass Prairie Site

Abstract:

Climatic variability influence vegetation condition and growth, which are often characterized by satellite derived vegetation indices (VIs) such as normalized difference vegetation Index (NDVI), enhanced vegetation Index (EVI) and land surface water index (LSWI). A fourteen-year (2000 – 2013) of NDVI, EVI and LSWI data from Moderate-Resolution Imaging Spectroradiometer (MODIS) were analyzed for assessing grassland vegetation dynamics to climatic variability ( i.e., drought) and soil moisture over the time series at the Marena site, Stillwater Oklahoma,USA. This is also the Marena, Oklahoma In-situ Sensor Testbed (MOISST) site in support of NASA SMAP mission. Change in magnitudes of VIs provided the dynamics of inter-annual variability of the grassland vegetation. The magnitude of VIs declined in dry years (2006 and 2012) .Variation in NDVI and EVI over years resulting from climatic variability was strongly correlated to cumulative seasonal rainfall (NDVI, r = 0.82, EVI, r = 0.77) and average seasonal soil volumetric water content up to 60 cm depth. The EVI declined more than did NDVI during the dry events of 2006 and 2012, indicating that the use of EVI in place of NDVI appears to increase the performance of ecosystem models under drought condition. Furthermore, LSWI was the most sensitive index to drought among the three VIs tested. The LSWI values were negative (LSWI < 0) even within the middle of plant growing season in dry years, showing its potential to track the hydrological status of the ecosystem. Duration of LSWI < 0 in summer was greater in dry years (2006= 53 & 2012= 34 days) compared to wet years (2007 & 2013= 0 days). The result also revealed that LSWI values quantitatively corresponded well with the drought severity categories identified by the United States Drought Monitoring (USDM). In conclusion, the number of days with LSWI < 0 during the plant growing season and LSWI-based drought severity classification scheme can provide useful information for an assessment of the drought impacts over grasslands.