Evapotranspiration Estimation with the S-SEBI Method from Landsat 8 Data against Lysimeter Measurements at the Barrax Site, Spain

Evapotranspiration (ET) is a variable of the climatic system and hydrological cycle that plays an important role in biosphere-atmosphere-hydrosphere interactions. In this paper, remote sensing-based ET estimates with the simplified surface energy balance index (S-SEBI) model using Landsat 8 data were compared with in situ lysimeter measurements for different land covers (Grass, Wheat, Barley, and Vineyard) at the Barrax site, Spain, for the period 2014-2018. Daily estimates produced superior performance than hourly estimates in all the land covers, with an average difference of 12% and 15% for daily and hourly ET estimates, respectively. Grass and Vineyard showed the best performance, with an RMSE of 0.10 mm/h and 0.09 mm/h and 1.11 mm/day and 0.63 mm/day, respectively. Thus, the S-SEBI model is able to retrieve ET from Landsat 8 data with an average RMSE for daily ET of 0.86 mm/day. Some model uncertainties were also analyzed, and we concluded that the overpass of the Landsat missions represents neither the maximum daily ET nor the average daily ET, which contributes to an increase in errors in the estimated ET. However, the S-SEBI model can be used to operationally retrieve ET from agriculture sites with good accuracy and sufficient variation between pixels, thus being a suitable option to be adopted into operational ET remote sensing programs for irrigation scheduling or other purposes.

Automated summary

The study compared remote sensing-based ET estimates using the S-SEBI model with in situ lysimeter measurements for different land covers in Barrax, Spain from 2014 to 2018. Daily estimates outperformed hourly estimates, with Grass and Vineyard showing the best performance. The S-SEBI model demonstrated the ability to retrieve ET from Landsat 8 data with good accuracy and variability between pixels, making it a suitable option for operational ET remote sensing programs.