Soil temperature at ECMWF: An assessment using ground‐based observations. Issue 4 (26th February 2015)

• Record Type: Journal Article
• Authors: Albergel, C.; Dutra, E.; Muñoz‐Sabater, J.; Haiden, T.; Balsamo, G.; Beljaars, A.; Isaksen, L.; de Rosnay, P.; Sandu, I.; Wedi, N.
• Is Part Of: Journal of geophysical research. Volume 120:Issue 4(2015:Mar.)
• Page Start: 1361
• Page End: 1373
• Publication Date: 2015-02-26
• Subjects: Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1002/2014JD022505
• Languages: English
• ISSNs: 2169-897X

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>Soil temperature is an important variable for the representation of many physical processes in numerical weather prediction (NWP). It is the key driver for all surface emissions of energy, carbon dioxide, and water and forward operator for all satellite sensors sensitive to land. Yet the forecast quality of this variable in NWP is largely unknown. In this study, in situ soil temperature measurements from nearly 700 stations belonging to four networks across the United States and Europe are used to assess the European Centre for Medium‐Range Weather Forecasts (ECMWF) forecasts of soil temperature during 2012. Evaluation of the time series shows a good performance of the short‐range forecasts (day one) in capturing both soil temperature annual and diurnal cycles with very high level of correlation (0.92 and over), averaged root‐mean‐square differences ranging from 2.54°C to 3.89°C and averaged biases ranging from −0.52°C to 0.94°C. The orography data set used in the forecast system was found to have a strong impact on the outcomes of the evaluation. The difference between elevation of a station and that of the corresponding grid cell in the ECMWF model may lead to large temperature differences linked to linear processes resulting in a constant bias, as well as nonlinear processes (e.g., to snow melt in spring). This verification study aims to contribute to a better understanding of the near‐surface forecasts performance highlighting land‐atmosphere processes that need to be better represented in future model development such as snow pack melting and heat diffusion in the soil.</p> </abstract>