Scholarlink

<p>Remote sensing-based evapotranspiration (ET) estimation is essential for hydrological and agricultural applications but is strongly constrained by cloud-induced gaps in thermal infrared land surface temperature (LST) observations. Although all-weather, gap-filled LST products have demonstrated good radiometric accuracy, their impact on ET estimation, particularly at aggregated timescales relevant for water management applications remains poorly quantified. This study evaluates how enhanced LST temporal continuity influences ET estimates derived from a physically based surface energy balance model. ET is estimated using the Soil Plant Atmosphere and Remote Sensing Evapotranspiration (SPARSE) Patch model driven by clear-sky MODIS LST and a passive microwave–based all-weather LST product and is evaluated against eddy covariance measurements across seven sites in India at daily, weekly, and monthly timescales. At the daily scale, ET estimates driven by MODIS and passive microwave–based LST show comparable accuracy (RMSE ≈ 1.2&#xa0;mm day<sup>–1</sup>), confirming the physical consistency of passive microwave thermal information. In contrast, aggregation reveals clear advantages of enhanced LST temporal continuity. Weekly all-weather ET reduces aggregate RMSE by approximately 13% relative to MODIS-only ET while increasing data availability by 53%. The benefits are consistent at the monthly scale, where all-weather ET reduces RMSE by 12% and increases the number of valid observations by 50%. These results demonstrate that temporal sampling limitations of clear-sky LST are a major source of uncertainty in aggregated ET estimates and that all-weather LST gap filling substantially improves ET reliability at management-relevant timescales.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The utility of all-weather land surface temperature for energy balance–based evapotranspiration estimation

  • 期刊论文

  • 发表日期：

    2026年4月24日

查看全文
  • Rahul Harod,
  • K. V. Athira,
  • Eswar Rajasekaran

摘要

Remote sensing-based evapotranspiration (ET) estimation is essential for hydrological and agricultural applications but is strongly constrained by cloud-induced gaps in thermal infrared land surface temperature (LST) observations. Although all-weather, gap-filled LST products have demonstrated good radiometric accuracy, their impact on ET estimation, particularly at aggregated timescales relevant for water management applications remains poorly quantified. This study evaluates how enhanced LST temporal continuity influences ET estimates derived from a physically based surface energy balance model. ET is estimated using the Soil Plant Atmosphere and Remote Sensing Evapotranspiration (SPARSE) Patch model driven by clear-sky MODIS LST and a passive microwave–based all-weather LST product and is evaluated against eddy covariance measurements across seven sites in India at daily, weekly, and monthly timescales. At the daily scale, ET estimates driven by MODIS and passive microwave–based LST show comparable accuracy (RMSE ≈ 1.2 mm day–1), confirming the physical consistency of passive microwave thermal information. In contrast, aggregation reveals clear advantages of enhanced LST temporal continuity. Weekly all-weather ET reduces aggregate RMSE by approximately 13% relative to MODIS-only ET while increasing data availability by 53%. The benefits are consistent at the monthly scale, where all-weather ET reduces RMSE by 12% and increases the number of valid observations by 50%. These results demonstrate that temporal sampling limitations of clear-sky LST are a major source of uncertainty in aggregated ET estimates and that all-weather LST gap filling substantially improves ET reliability at management-relevant timescales.