Advanced Search
ZHU Yanliang, LING Chao, CHEN Hongbin, ZHANG Jinqiang, PENG Liang, YU Yu. Comparison of Two Reanalysis Data with the RS92 Radiosonde Data[J]. Climatic and Environmental Research, 2012, 17(3): 381-391. DOI: 10.3878/j.issn.1006-9585.2011.10138
Citation: ZHU Yanliang, LING Chao, CHEN Hongbin, ZHANG Jinqiang, PENG Liang, YU Yu. Comparison of Two Reanalysis Data with the RS92 Radiosonde Data[J]. Climatic and Environmental Research, 2012, 17(3): 381-391. DOI: 10.3878/j.issn.1006-9585.2011.10138

Comparison of Two Reanalysis Data with the RS92 Radiosonde Data

  • An intercomparison by interpolation is made between two types of reanalysis data (NCEP/NCAR and ERA-Interim) and seven months (May 2008 to December 2008) measurements of high vertical resolution GPS radiosonde (RS92) at Shouxian station (32.56°N, 116.78°E), China. The correlation coefficient (R), bias (Bias), and mean absolute deviation (Mad)of temperature (T), zonal wind (U), meridional wind (V), and relative humidity (RH) are presented at the standard pressure levels of two reanalysis data sets from 1000 hPa to 20 hPa. The results indicate better correlation of RS92 observations with the ERA-Interim data than that with the NCEP/NCAR data. T, U, and V have higher correlation coefficients than RH at all levels. R of T is close to 1.0 from 1000 hPa to 250 hPa and decreases gradually above 250 hPa. The absolute values of bias for T is mostly less than 0.3 ℃ between ERA-Interim and RS92, but about two times between NCEP and RS92 at all levels. R of U is larger in middle and upper troposphere than in lower troposphere and lower stratosphere. R of V increases with height in the troposphere and decreases dramatically in lower stratosphere. R of RH decreases with height. It is shown that Bias of RH ranges from 10% to 20% through 400 hPa to 100 hPa and becomes smaller than 10% in the upper layers, which is in part attributed to the limitation of measurement precision.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return