Evapotranspiration, Gross Primary Production and Water-Use Efficiency Estimates of Cotton and Corn in East-Central Texas Using Eddy Covariance and Planetscope.

Accurate simulations of evapotranspiration (ET) and gross primary production (GPP) are important in quantifying the contributions of agricultural production to global carbon and hydrologic cycles. The objectives of this study were as follows: to model GPP using PlanetScope satellite remote sensing, to compare modeled GPP to GPP measured with eddy covariance, and to study the relationship between GPP and ET. Eddy covariance data was collected using two flux towers established in two near-by fields in Burleson County, TX in the 2017 and 2018 growing seasons; one planted to corn (Zea mays L.) and one to cotton (Gossypium hirsutum L.). These measurements were made using an open-path eddy covariance system consisting of a CSAT-3A sonic anemometer (Campbell Scientific Inc., Logan, UT, USA) and LI-7500A infrared gas analyzer (LI-COR Biosciences, Lincoln, NE, USA). Each tower had additional meteorological instruments, including photosynthetically active radiation (PAR) sensors. Leaf area index (LAI) and other phenological data was collected with biweekly destructive sampling. Cloud-free PlanetScope images were downloaded for both growing seasons. Band values were extracted and converted to top of atmosphere reflectance. The following vegetation indices (VI’s) were calculated: Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Weighted Difference Vegetation Index (WDVI), and Enhanced Vegetation Index (EVI2). A GPP model was developed where GPP » VI * PAR *LAI. The 2017 data was used for model calibration and the 2018 data was used for model validation. The corn GPP models performed adequately with the most accurate model being the SAVI-based model (nRMSE = 0.13 g C /day and r2 = 0.96). The cotton GPP models performed less accurately, with the most accurate cotton model being the NDVI-based model (nRMSE = 0.24 g C / day and r² = 0.87).