11月27日下午，陈学龙博士应邀到水保所做了题为A roughness height parameterization method for global land surface heat fluxes and evapotranspiration的报告

陈学龙, 博士, 2003年毕业于兰州大学大气科学系，后留校工作，2011年毕业于中科院青藏高原研究所，2013年毕业于荷兰Twente大学水资源系，现就职于Twente大学，近5年在ACP,HESS, JGR, JAS, QJRMS等国际刊物发表学术论文20篇, 担任RSE, JGR,HESS,JHM, JAMC等国际期刊审稿人, 研究方向为地气交换的湍流参数化及遥感的应用，能量水循环研究等。

Estimating the sensible heat flux (H) over vegetation from thermalinfrared temperature requires an estimate of the excess resistance kB^-1 (where k is von Karman constant and kB^-1 is the inverse Stantonnumber), the difference in turbulenttransfer efficiency between momentum and scalars. kB^-1has been the subject ofconsiderable interest in micrometeorology, land surface model, and surfaceturbulent fluxes simulations, but there still does not exist a uniform method for use in remote sensingretrieval of land surface flux. This study is motivated by the application ofone-dimensional turbulent diffusion methods to describe canopy-atmosphereinteraction with remote sensing global land surface variables. Firstly, theuncertainties of the estimated sensible heat flux due to kB^-1 parameterization for 7 different land covers(including needle forest, broadleaf forest, shrub, savanna, grassland,cropland, and sparsely vegetated land) were examined. We then tested andverified the performance of the kB^-1  scheme in Su 2002 (in short Su02) bycomparing H estimated by Su02 and H measured at 28 flux tower stations. Large differences in the accuracy of thesimulated H are found for the seven land covers. The model predictionsof H for grass, crop and sparsely vegetated land compare favorably withobserved values. H is significantlyunderestimated at forest sites (whether needle or broadleaf) due to its higher kB^-1 estimate for the canopy part. Several solutions and newmethod were innovated to eliminate thelower estimation. An effectiveleaf drag coefficient, effective leaf heat transfer coefficients,foliage shelter factor and new within-canopy momentum model was used to reduceunderestimates of turbulent heat flux for high canopy because the shelteringeffects and variable drag coefficients are not included in Su02 kB^-1 scheme. The model predictions of H for grass, crop and sparsely vegetatedland compare favorably with observed values, when the canopy information isgiven an actual value. H issignificantly underestimated at forest sites due to a high value of kB^-1. Among thedifferent physical representation for the canopy, canopy-soil mixture, and soilcomponent, it is found that such high  kB^-1 value is caused by the high kB^-1 value for the canopy part. The reasons for high kB^-1 was investigated from canopy-air physicalprocess of turbulent diffusion. Subsequently, this study introduces thevertical foliage density information into a column canopy-air turbulentdiffusion model toinclude the different momentum and heat transfer efficiency in the verticalcanopy layersto enhance the thermal turbulent transfer intensity above the high canopy. Thenew model has been verified to have high accuracy over different canopystructures. The newcanopy-atmosphere turbulent transfer model has been used to generate heat fluxmaps over the global land.