Improved agricultural Water management in data-scarce semi-arid watersheds: Value of integrating remotely sensed leaf area index in hydrological modeling

• Published in: The Science of the total environment Vol. 791; p. 148177
• Main Authors: Paul, Manashi, Rajib, Adnan, Negahban-Azar, Masoud, Shirmohammadi, Adel, Srivastava, Puneet
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.10.2021
• Subjects: Decision tree; Deficit irrigation; Leaf Area Index; Semi-arid regions; SWAT; Water productivity; SWAT; Deficit irrigation; Leaf Area Index; Semi-arid regions; Decision tree; Water productivity
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2021.148177

In watersheds located in semi-arid regions, vegetation dynamics, evapotranspiration (ET), and associated water and energy balances collectively play a major role in controlling hydrological regimes and crop yield. As such, it is challenging to predict the complex hydrological processes and biophysical dynamics. This challenge increases in areas with limited data availability. The key objective of this study was to evaluate the direct integration of remotely sensed Leaf Area Index (LAI) data into a hydrological model to improve streamflow, ET, and crop yield estimates. We also demonstrated how an improved model integrated with remotely sensed LAI data can inform water managers by predicting water productivity (WP) under different irrigation schemes. We took agricultural-dominated San Joaquin Watershed in California, United States, as our testbed and integrated the Moderate Resolution Imaging Spectroradiometer (MODIS) 500-m resolution 4-day total LAI data into the SWAT (Soil and Water Assessment Tool) model. Results showed that, compared to conventional SWAT model that relies on semi-empirical equations and user inputs for simulating biophysical processes, direct LAI integration into SWAT model (SWAT-LAI) notably captured the actual vegetation dynamics and improved ET and crop yield estimations. The WP simulated by the improved SWAT-LAI model for almond and grape yields varied within a range from 0.363 to 3.81 kg/m3 and 0.32 to 4.76 kg/m3 across different irrigation applications. The outcomes of this study showed that deficit irrigation application could be a viable option in water stressed regions, since it can save a substantial amount of irrigation water and maintain the higher water productivity required for both almond and grape yield production. This study shows an evidence of how remotely sensed data integrated into hydrological models can serve as a decision support tool by providing quantitative information on crop water use and crop production. [Display omitted] •Integrated remote sensing LAI data into the hydrological model improved ET and streamflow simulations.•Crop yield estimation was enhanced when high-resolution MODIS LAI data inserted into the SWAT model.•Auto or precise irrigation applications ensure more water use efficiency for grape by increasing the WP.•Deficit irrigation application can save a substantial amount of irrigation water while maintaining higher water productivity for almonds.