Open AccessDownscaling of ERA-Interim Temperature in the Contiguous United States and Its Implications for Rain–Snow Partitioning
Author(s) -
Guoqiang Tang,
Ali Behrangi,
Zhanshan Ma,
Di Long,
Yang Hong
Publication year - 2018
Publication title -
journal of hydrometeorology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.733
H-Index - 123
eISSN - 1525-755X
pISSN - 1525-7541
DOI - 10.1175/jhm-d-18-0041.1
Subject(s) - downscaling , environmental science , climatology , snow , precipitation , meteorology , geopotential height , global precipitation measurement , geology , geography
Precipitation phase has an important influence on hydrological processes. The Integrated Multisatellite Retrievals for Global Precipitation Measurement (IMERG) uses temperature data from reanalysis products to implement rain–snow classification. However, the coarse resolution of reanalysis data may not reveal the spatiotemporal variabilities of temperature, necessitating appropriate downscaling methods. This study compares the performance of eight air temperature Ta downscaling methods in the contiguous United States and six mountain ranges using temperature from the Parameter-Elevation Regressions on Independent Slopes Model (PRISM) as the benchmark. ERA-Interim Ta is downscaled from the original 0.75° to 0.1°. The results suggest that the two purely statistical downscaling methods [nearest neighbor (NN) and bilinear interpolation (BI)] show similar performance with each other. The five downscaling methods based on the free-air temperature lapse rate (TLR), which is calculated using temperature and geopotential heights at different pressure levels, notably improves the accuracy of Ta. The improvement is particularly obvious in mountainous regions. We further calculated wet-bulb temperature Tw, for rain–snow classification, using Ta and dewpoint temperature from ERA-Interim and PRISM. TLR-based downscaling methods result in more accurate Tw compared to NN and BI in the western United States, whereas the improvement is limited in the eastern United States. Rain–snow partitioning is conducted using a critical threshold of Tw with Snow Data Assimilation System (SNODAS) snowfall data serving as the benchmark. ERA-Interim-based Tw using TLR downscaling methods is better than that using NN/BI and IMERG precipitation phase. In conclusion, TLR-based downscaling methods show promising prospects in acquiring high-quality Ta and Tw with high resolution and improving rain–snow partitioning, particularly in mountainous regions.