Validation, Calibration, and Improvement of Remote Sensing ET Algorithms in Mountainous Regions Using Scintillometers.

摘要

The overall goal of this study was to improve the remote sensing estimates of evapotranspiration in mountainous regions by field studies at New Mexico sites that are representative of large parts of the mountainous regions in the arid southwest. In particular, we proposed to validate sensible and latent heat fluxes estimated from the remote sensing Surface Energy Balance Algorithms for Land (SEBAL) model using large aperture scintillometer (LAS) ground measurements over mountainous landscapes in New Mexico. Unfortunately, research is not always predictable. In our study, shortly after the award of the grant we suspected and later confirmed with inter-comparison measurements that the sensible heat flux measurements by the ten large aperture Kipp & Zonen scintillometers of Hendrickx’s research group contained a considerable inter-instrument error. Therefore, we had no choice but to shift the focus of the study toward (1) conducting a rigorous quantitative inter-comparison study for Kipp & Zonen large aperture scintillometers and (2) testing whether the biased scintillometers can still be used for field energy balance studies. Our conclusions are: (1) the first generation Kipp & Zonen LASs does indeed have a large inter-instrument bias that can exceed 21%; (2) these LASs can be used for field energy balance studies after calibration against a reference LAS or eddy covariance sensible heat flux measurements over homogeneous terrain. Our inter-comparison study by Kleissl et al. [2008] has been cited 56 times by studies reconfirming our findings or investigating error causes and remediation. More importantly the manufacturer Kipp & Zonen has redesigned its LAS and has started the production of the next generation LAS without inter-instrument biases. The New Mexico Scintillometer Network has taught us many valuable lessons on how to utilize scintillometry [Kleissl et al., 2009a] for statewide validation of evapotranspiration remote sensing algorithms in New Mexico.