Spatio-temporal dynamics of water use efficiency over forest ecosystems using time series satellite data and carbon flux measurements

Abstract

The significance of forests in limiting the consequences of climate change and maintaining the biogeochemical cycle is immense. Gross primary productivity (GPP) and evapotranspiration (ET) are the two most important measures of carbon and water fluxes in the forest ecosystem. The ratio between these two parameters is termed water use efficiency (WUE). In this study, the WUE was estimated for two contrasting forest ecosystems: a moist deciduous forest at the Barkot Flux Site (BFS) and a young mixed dry deciduous forest plantation at the Haldwani Flux Site (HFS), Uttarakhand, India. The Vegetation Photosynthesis Model (VPM) was utilized to estimate GPP, whilst the Mapping Evapo-Transpiration at High Resolution with Internal Calibration (METRIC) approach was employed to evaluate ET using Landsat 8 OLI satellite data from 2016 to 2018. Concurrent measurements of WUE as the ratio of GPP to ET, obtained from eddy covariance (EC) towers, were used for validation purpose. For both the sites, the maximum and minimum GPP were reported in October and December-January, respectively, whereas, the months of April-May and December had the highest and lowest ET, respectively. WUE remained elevated during November-December and reached its lowest point in April-May at both sites. The METRIC model estimated ET and EC-based GPP had reasonably good agreement with R2 values of 0.76 and 0.71, respectively. Consequently, the simulated WUE also matched closely with the observed WUE (R2 = 0.62). Accurate quantification of spatial WUE using the GPP and ET obtained from remote sensing-based VPM and METRIC models provided insight into its variability in different vegetation compositions. Further, the availability of long-term satellite data may pave the way for an improved understanding of forest ecosystems' response to water limitation in changing climate.