Estimating Biomass and Yield Using METRIC Evapotranspiration and Simple Growth Algorithms

Abstract

Core Ideas Crop models have significant limitations in estimating “real‐world” yields across the landscape. Joining satellite‐based evapotranspiration and leaf area index with growth algorithms provide good yield estimation. Proposed method evapotranspiration and yields compared well with observations at four dryland sites. Crop models are used to assess crop yield under prescribed scenarios and at scales varying from point to field to region and beyond. The use of models to evaluate the performance of agricultural systems, as they exist in the real world, can be challenging and plagued with constraints. This is due to the difficulty in characterizing the spatial variability across the landscape of crops, soils, weather, management, miscellaneous stress factors, and the initial state of the system. We propose the use of actual evapotranspiration (ETa) estimated from remote sensing images and simple crop growth‐transpiration algorithms as an alternative to the use of standalone crop models for real‐world yield assessment. In this study, we combined ETa estimates from METRIC (Mapping Evapotranspiration at High Resolution with Internalized Calibration) with simple crop growth algorithms extracted from the CropSyst model to estimate biomass production and yield at high resolution (30 by 30 m). We tested this approach in four dryland agriculture sites in eastern Washington State with contrasting annual precipitation. All sites were equipped with an eddy covariance flux tower for ground ETa estimation. The proposed approach was able to provide good estimates of ETa, seasonal change of aboveground biomass and yields at all sites when compared with observations for a 3‐year period, collectively including five different annual crops. Because estimations are made at high resolution, they can be scaled up to field or regional scales. Advantages and limitations of the proposed approach are discussed.