Homogenization of surface temperature data in High Mountain Asia through comparison of reanalysis data and station observations

摘要

High-quality temperature estimates with good spatio-temporal coverage are necessary for completely understanding the influences of warming climate on cryosphere and hydrological systems in High Mountain Asia (HMA). In this study, we compare reanalysis temperature data from ERA-Interim and National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) with station observations in HMA during 1979-2011. The results indicate that although reanalysis temperature data can capture the warming trends in HMA, the trend magnitudes are clearly underestimated by reanalysis data. In particular, the increase in summer temperature cannot be identified from the NCEP/NCAR reanalysis. For ERA-Interim, temperature increases are underestimated in the north and northwest of HMA; for NCEP/NCAR, the warming magnitudes show evident biases in the Pamir, Himalayas, and southeastern Tibetan Plateau. Considering that high-frequency signals and periodical fluctuations among the three datasets are in good agreement, and based on the wavelet transform method, the low-frequency component decomposed from the temperature time series of ERA-Interim and NCEP/NCAR reanalyses is adjusted by that derived from station observations. The resulting homogenized reanalysis temperature data show much better spatio-temporal consistency with station data. The differences in monthly and annual temperature anomalies between station and homogenized ERA-Interim and NCEP/NCAR reanalysis data become more convergent. The homogenized temperature time series are better correlated with station data at annual and seasonal timescales.