Satellite imagery provides an unprecedented spatial distribution of critical land surface parameters.Numerous physical and empirical remote sensing-based models

in conjunction with ancillary surface and atmospheric data

have been developed to estimate evapotranspiration（ET） in the quantitative thermal remote sensing field.ETWatch is an operational scheme for watershed ET monitoring

which is designed to integrate with various types of remote-sensed ET models to obtain continuous ET maps.To reduce the errors from modeling

we improved ETWatch to carry out calibration on key variables such as surface temperature

ground temperature difference

short-wave and long-wave radiation

water heat flux

and sensible heat flux

using ground measurement data.The validation results show that the calibration greatly improved the accuracy of ET products

and the importance of calibration cannot be overemphasized in applications used in watershed water management.