格点尺度对TRMM微波成像仪云水数据的影响

Impact of Gridding Scale on TRMM Microwave Imager Cloud Water Information

选取热带测雨卫星(Tropical Rainfall Measuring Mission,TRMM)微波成像仪(TRMM Microwave Imager,TMI)液态水路径(liquid water path,LWP)轨道像元数据为研究对象,探讨了将瞬时探测以及逐月的像元数据进行格点化(0.1°、0.25°、0.5°、1.0°和2.5°五种格点分辨率)时,格点数据的失真情况。对TMI瞬时探测的个例分析结果表明,细分辨率(0.1°、0.25°和0.5°)格点能保留原始像元数据的细节;而随着网格变粗,细节受到较大的平滑。因此对于中尺度到天气尺度的天气系统分析而言,将卫星轨道数据处理到网格尺度不大于0.5°的格点更合适。对逐月LWP像元资料格点化处理的分析表明,细分辨率格点能保留LWP空间分布细节,尽管5种分辨率下LWP的概率密度分布(probability density function,PDF)均相近。因此,对月尺度及以上的气候分析研究而言,格点尺度大小对卫星像元数据格点化的影响不显著。最后利用本实验室计算的TMI/LWP格点数据与欧洲中期数值预报中心再分析资料(European Centre for Medium-range Weather Forecasts Interim reanalysis,ERA-Interim)和NCEP再分析资料(NCEP Climate Forecast System Reanalysis,NCEP CFSR)进行了对比,发现两种再分析资料都高估了LWP;TMI/LWP格点数据与两种再分析资料LWP的多年变化趋势大致相同。

Satellite observation data are valuable for model evaluation, but to readily compare satellite-based data with model simulations, swath pixel data must first be gridded. In this study, data distortions caused by gridding （0.1°, 0.25°, 0.5°, 1.0°, and 2.5° gridding resolutions） are investigated using instantaneous and monthly pixel data from the Tropical Rainfall Measuring Mission （TRMM） Microwave Imager （TMI） liquid water path （LWP） data. Results from this case study show that data gridded at grid scales of 0.1°, 0.25°, and 0.5° retain more local details in the instantaneous pixel data, while the details tend to be smoothed out at coarser grid resolutions. So data with grid scales no coarser than 0.5° are suitable for analyzing weather activities from the mesoscale to the synoptic scale. In terms of the monthly pixel LWP, data with gridding resolutions of 0.1°, 0.25°, and 0.5° also retain more detail. Although, probability density functions （PDF） of the LWP show similar patterns at all the gridding scales considered. So when analyzing monthly LWP data, the impact of gridding scales is not significant. Finally, this study compared the gridded TMI LWP data with the European Centre for Medium-range Weather Forecasts Interim reanalysis （ERA-Interim） and NCEP Climate Forecast System Reanalysis （CFSR）, and found that both of these reanalyses overestimate the magnitudes of the LWP. Gridded TMI LWP data, the ERA-Interim, and the CFSR show reasonably similar regularities in the variance of the LWP.