Environmental Lapse Rate for High‐Resolution Land Surface Downscaling: An Application to ERA5

• Published in: Earth and space science (Hoboken, N.J.) N.J.), 2020-05, Vol.7 (5), p.n/a
• Main Authors: Dutra, Emanuel, Muñoz‐Sabater, Joaquín, Boussetta, Souhail, Komori, Takuya, Hirahara, Shoji, Balsamo, Gianpaolo
• Format: Article
• Language: English
• Subjects: Air temperature; Climate change; Climatology; Data collection; downscaling; Humidity; land surface; lapse rate; Precipitation; Rain; reanalysis; Simulation; Snow; Snow depth; Soil temperature; Topography; Troposphere; Weather forecasting
• ISSN: 2333-5084; 2333-5084
• DOI: 10.1029/2019EA000984

In this study we derive the environmental lapse rate (ELR) from vertical profiles of temperature in the lower troposphere, applying it to downscale air temperature of the new European Centre For Medium‐Range Weather Forecasts (ECMWF) reanalysis ERA5, which replaces ERA‐Interim (ERAI). We focus over the western U.S. region, a data‐rich area with observations of daily maximum and minimum temperature (Global Historical Climatology Network) and snow depth and soil temperature. Observations indicate an ELR of −4.5 K·km−1 in the region, lower than the commonly used −6.5 K·km−1. ERA5 ELR agrees with the observational estimates, with some overestimation in winter and limitations in the diurnal variability. The elevation correction of ERA5 temperature using different ELR showed the benefits of deriving ELR fields from ERA5 vertical profiles, when compared with a constant ELR. Simulations with the ECMWF land surface model, at 9‐km resolution, driven by ERA5 using different ELR corrections showed the added value of the methodology, but the impact of different ELR corrections is limited. However, the validity of the downscaling method in reducing temperature to station altitude suggests that there is sufficient generality for application at kilometer and subkilometer resolutions. By comparing the estimated representativity errors of observations with reanalysis, the improvements from ERAI to ERA5 are mainly visible in the random component of the error. Large systematic biases remain, which require further attention from the modeling and data assimilation, and limit the potential benefits of ELR corrections. Key Points Environmental lapse rate derived from atmospheric reanalysis vertical profiles agrees with observational estimates Surface downscaling outperforms ERA5, but the impact of different ELR corrections to the driving data is reduced Systematic biases in ERA5 near‐surface temperature require further efforts from modeling and data assimilation