Evaluation of the SEBAL for Estimating Actual Evapotranspiration in the Area of the Gareh Bygone Plain Floodwater Spreading Project

Abstract

Introduction

Accurate estimation of actual evapotranspiration (ETa) is essential for sustainable water resource management, particularly in arid and semi-arid regions. This study employed the SEBAL model in conjunction with Landsat 8 and 9 satellite imagery to estimate ETa in the Gareh Bygone Plain, Fars Province, during 2018–2021. The model integrated corrected satellite data and meteorological inputs to compute key surface energy balance components, including net radiation, soil heat flux, and latent heat flux. Calibration was enhanced by incorporating wind speed data, improving the model’s accuracy. ETa values varied seasonally, ranging from 0.8 to 3.2 mm/day in colder months and 1.8 to 6.5 mm/day during warmer periods. The model’s results were validated against FAO Penman-Monteith reference ET and field measurements, confirming strong agreement. Crop coefficient (Kc) estimates highlighted significant variability based on vegetation type and growth stages. The complementary use of the SEBS model yielded ETa estimations with a low error margin (3–5%), further confirming the reliability of remote sensing-based approaches. These findings support the application of calibrated satellite models and localized parameters in optimizing irrigation strategies and addressing water scarcity in semi-arid environments.

Accurate estimation of actual evapotranspiration (ETa) is crucial for water resource management and evaluating the efficiency of artificial recharge projects in arid regions. Traditional methods relying on point-based measurements often fail to represent large-scale and heterogeneous areas. Remote sensing technologies, utilizing satellite data and surface energy balance models such as SEBAL, enable precise estimation of ETa and crop coefficients (Kc) over extensive spatial and temporal scales. This study aimed to assess water consumption of various vegetation covers and the effectiveness of the flood spreading system in the Gareh Bygone plain, Fars Province, by developing improved models and analyzing the spatiotemporal distribution of ETa.



Materials and Methods

This study was carried out at the Kosar floodwater spreading station in the semi-arid Gareh Bygone Plain of southern Iran’s Fars Province.The Gareh Bygone Plain, with an approximate area of 192 square kilometers, is located downstream of three main watershed areas. The area features an semi-arid to arid climate, an average annual precipitation of 229 mm, and geology comprising limestone, marl, and conglomerates. Satellite data from Landsat 8 and 9, processed with geometric, radiometric, and atmospheric corrections, served as input for the SEBAL model. This model estimates actual evapotranspiration by integrating satellite imagery and meteorological data using the surface energy balance equation. Vegetation indices such as NDVI and SAVI, surface albedo, surface temperature, and other surface energy parameters were used in ETa estimation. Model results were validated with field measurements, including reference evapotranspiration, soil water balance, and estimates of return flow.



Results and Discussion

Satellite imagery from Landsat 8 Level 2, covering 2018–2021, was used to derive ETa maps via the SEBAL model. Due to data acquisition limitations in certain provinces, temporal coverage was incomplete. The SEBAL model, calibrated with meteorological data and incorporating wind speed as a parameter, showed good performance in estimating ETa, particularly in the Gareh Bygone plain. Comparison of SEBAL-derived ETa values with FAO Penman-Monteith reference estimates, using a crop coefficient (Kc=1.05), demonstrated the model’s reasonable accuracy. ETa values ranged from 0.8 to 3.2 mm/day during the cold season and from 1.8 to 6.5 mm/day in the warm season, reflecting a seasonal pattern consistent with the region’s climatic conditions. These results highlight the importance of using satellite data and optimizing model parameters for more accurate ETa estimation. Analysis of SEBAL model results revealed seasonal fluctuations in ETa influenced by temperature, solar radiation, and vegetation cover. ETa values ranged from 0.8 to 6.5 mm/day, peaking during the warm season. The crop coefficient (Kc) was closely linked to vegetation type and growth stage, emphasizing the need to consider these differences in water management. Field measurements of soil (moisture, texture, bulk density) in the Gareh Bygone plain indicated a progressive increase in water infiltration depth and return flow volume during the irrigation season. Comparison of SEBS model-derived ETa with field data confirmed the high accuracy of the SEBS model (error margin of 3–5%), demonstrating its superiority over traditional methods and its effectiveness for agricultural water resource management in semi-arid regions. Computational results showed an increasing trend in return flow volume throughout the 11 irrigation cycles, reaching a significant amount by the end of the season. This finding has important implications for water resource management in the region.



Conclusion

This study employed Landsat 8 imagery to estimate actual evapotranspiration (ETa) using the SEBAL model in the semi-arid Gareh Bygone Plain. Results indicate a strong correlation between seasonal ETa variations and temperature, solar radiation, and land surfaces vegetation activity, with peak ETa during the warm season (April-October) and minimum during the cold season (December-February). This seasonal pattern aligns with the semi-arid climate and highlights the importance of considering spatiotemporal variability in ETa estimations. Furthermore, incorporating wind speed significantly improved SEBAL model accuracy.