An Upscaling Analysis on Aerodynamic Roughness Length in North Tibetan Plateau based on Eddy Covariance, Large Aperture Scitillometer Data and Remote Sensing Product

Abstract:

Abstract: Aerodynamic roughness length (z0m) is a crucial parameter in quantifying momentum, sensible heat, and latent heat fluxes between atmosphere and land surface, and depends greatly on spatial scales. This paper presents a tentative study in upscaling of z0m in North Tibetan Plateau, based on ground measurement data of different spatial scales from eddy covariance (EC) and large aperture scintillometer (LAS) and NDVI products from MODIS with 250m and 2km spatial resolutions. The comparison of z0m calculated from EC and LAS data indicates that the z0m values at both scales have apparent seasonal variations, and are in good agreement with that of NDVI. However, z0m_LAS is higher than z0m_EC, which is attributed to the differences of roughness elements in their footprints. An upscaling relationship about z0m was established with z0m observations and NDVI products of MODIS.In addition, an altitude correction factor was introduced into vegetation height estimation with NDVI, because the low temperature environment in North Tibetan Plateau due to its high altitude has strong influence on vegetation heights.The z0m retrievals with NDVI products with 250m spatial resolutions from June to September are validated with ground z0m results of Naqu/Amdo, Naqu/MS3478 and Naqu/NewD66 and the agreement is acceptable. The spatial distribution of z0m at 250m spatial resolutions in North Tibetan Plateau from June to September shows that z0m values are below 0.015m in most area, except the area in the southeast part where z0m values reach 0.025m due to lower altitudes. The z0m retrievals at 2km spatial resolutions of the same period range from 0.015 to 0.065m, and high values appear in the area with lower altitudes.The z0m retrievals at both spatial scales are affected by altitude, indicating the uniqueness of Tibetan Plateau. Frequency statistics on z0m retrievals at both spatial resolutions from June to September, 2012 shows obvious typical monthly changes in monsoon season.