Abstract
AbstractAccurate hourly two‐metre temperature gridded fields available in near real‐time are valuable products for numerous applications, such as civil protection and energy production planning. An analysis ensemble of temperature is obtained from the combination of a numerical weather prediction ensemble (background) and in situ observations. At the core of the flow‐dependent spatial interpolation method lies the analysis step of the local ensemble transform Kalman filter (LETKF). A scaling factor and a localization procedure have been added to correct for deficiencies of the background. Each observation is characterized by its own representativeness, which is allowed to vary in time. We call the method described here an Ensemble‐based Statistical Interpolation (EnSI) scheme for spatial analysis and it has been integrated into the operational post‐processing systems in use at the Norwegian Meteorological Institute (MET Norway). The benefits of the analysis are assessed over a 1‐year time period (July 2017–July 2018) and a case‐study is presented for a challenging situation over complex terrain. EnSI gives more accurate results than an interpolation method based exclusively on observations. The analysis ensemble provides a more informative representation of the uncertainty than a spatial analysis based on a single‐field background. EnSI reduces the number of large prediction errors in the analysis compared to the background by almost 50%, reduces the ensemble spread and increases its accuracy.
Highlights
Near-surface fields of temperature are among the most widely used products made available by national meteorological services
As pointed out by Amezcua and Leeuwen (2014), the analysis step of the Kalman filter is optimal when “(a) the distribution of the background is Gaussian, (b) state variables and observations are related via a linear operator, and (c) the observational error is of additive nature and has a Gaussian distribution”
Ensemble-based Statistical Interpolation (EnSI) has been compared against two optimal interpolation (OI)-based analysis schemes: (a) an observations-only OI which makes no use of Numerical weather prediction (NWP) model output, and (b) a “classical” OI using a single field as the background, instead of an ensemble
Summary
Near-surface fields of temperature are among the most widely used products made available by national meteorological services. As pointed out by Amezcua and Leeuwen (2014), the analysis step of the Kalman filter is optimal when “(a) the distribution of the background is Gaussian, (b) state variables and observations are related via a linear operator, and (c) the observational error is of additive nature and has a Gaussian distribution” These conditions are generally considered valid for two-metre temperature, as the widespread application of statistical techniques based on similar assumptions demonstrate (Haylock et al, 2008; Frei, 2014). Because of the limited number of ensemble members, Pb may contain spurious long-distance correlations and to deal with this issue we have: (a) implemented a gridpoint-by-gridpoint analysis scheme based on Hunt et al (2007), where for each grid point a local domain is defined and only the nearest observations are considered, and (b) introduced an R localization technique (Greybush et al, 2011), where the R elements are multiplied by a distance-dependent function. Note that the localization function ρ in Equation 11 decreases near the boundary of the local domain, the observation error variance is increased
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
