Satellite measurements of changes in water storage and their impact on vegetation dynamics

Estimating dynamics of groundwater at regional or global scale is difficult because we cannot directly measure the amount of water under the ground. Currently, ten years data of GRACE equivalent water thickness data are available to study global scale groundwater dynamics. GRACE data is useful to understand change in water storage at global scale, but the mechanisms of the water storage dynamics and its impacts on vegetation are not well understood. In this study, we tried to explain the interannual variations in the GRACE equivalent water thickness in relation to other satellite data, such as TRMM precipitation and MODIS land surface temperature. Then, we assessed the effect of dynamics of ground water on vegetation at both regional and global scale. At first, we decomposed the GRACE equivalent water thickness record into year-to-year variations and ten-year trend. The year-to-year variations in GRACE data are the result of annual water budget, while the ten-year trend in GRACE data can be explained by the trend in annual water storage and additional human water usage, especially irrigation. The year-to-year variation in GRACE data was spatially matched well with TRMM annual precipitation, but the high correlations with LST were found only in a few regions. The most of significant trend in GRACE data can be explained with long-term TRMM annual precipitation trend. The trend that cannot be explained by TRMM precipitation trend can be considered as the human water usage. Finally, we focused on the regional analysis in India to scrutinize the relationship among those satellite data including seasonal analysis. It is well known that irrigation in northwest India is exploiting groundwater and that was observed in GRACE data as decreasing trend in equivalent water thickness. On the other hand, western India showed increasing trend in NDVI for the last decades. TRMM data showed no significant trend in precipitation in India for the last decade. Those inconsistencies were examined in seasonal to interannual scales.