Search

This study reveals a close connection between interannual variation of the Somalia Jet (SMJ) intensity in boreal spring and the northwest-southeast movement of the South Asian High (SAH) in the following summer based on multiple datasets and numerical experiments. It is found that the summer SAH tends to shift northwestward (southeastward) when the preceding spring SMJ is stronger (weaker). There are two ways by which the spring SMJ intensity affects the following summer SAH. One is via modulating sea surface temperature (SST) in the tropical Indian Ocean and the other is through modulating anomalous heating associated with the Indian summer monsoon rainfall. On one hand, increase in the spring SMJ intensity enhances southwesterly monsoon winds and upward latent heat flux and leads to SST decrease in the tropical western Indian Ocean. The tropical western Indian Ocean cooling further leads to a decrease (increase) in upper-level geopotential height to the southeastern (northwestern) side of the SAH, inducing a northwestward movement of the SAH. On the other hand, increase in the spring SMJ intensity strengthens the Indian summer monsoon and enhances rainfall and atmospheric heating over the Indian subcontinent. The enhanced Indian subcontinent heating induces strong upper-level positive geopotential height to the northwestern side of the SAH, contributing to a northwestward shift of the summer SAH. Results of this study indicate that variation of the spring SMJ intensity is an effective predictor in the prediction of the northwest-southeast movement of the SAH in the following summer. [ABSTRACT FROM AUTHOR]

• Evaluate ET estimation uncertainty derived from remote sensing ET models. • Discuss ET estimation uncertainty derived from reanalysis products. • Investigate the effect of the uncertainty in eddy covariance tower observations. Since evapotranspiration (ET) is the intrinsic link between global energy and water cycle, remote sensing-based models have been developed for regional and global scale ET on heterogeneous land surface over the past four decades. In view of the significantly different physical mechanisms and mathematical expressions among remote sensing ET models as well as data availability and quality control process among the model input products, it is necessary to investigate the uncertainties of the multiple sources in actual ET estimation. Here, three remote sensing ET models, including the PT-DTsR model, the PM-mod model and the PML model, were simultaneously driven by three meteorological reanalysis products, resulting in nine calculation schemes to analyze the combined effect of the models and the input datasets. The Sobol' sensitivity method was also adopted for identifying the influential model parameters and in turn understanding the model process. The results indicated that estimates from nine calculation schemes showed great differences in the magnitude and temporal variation, explaining 20–50% of ET variability over all sites. Additionally, schemes compared with both uncorrected and corrected energy balance observations, as well as schemes using meteorological variables from three reanalysis products and Eddy Covariance tower observations, verified that the uncertainties in latent heat flux data observations caused by the energy budget mis-closure problem and spatial scale mismatch have propagated into the ET estimation. Our study is a beneficial reference for the uncertainties in remote sensing-based methods, and thus can provide guidance for the future development of ET models. [ABSTRACT FROM AUTHOR]