Efficient high-resolution 3-D interpolation of meteorological variables for operational use

Abstract

In recent years, the use of mesoscale meteorological network data has been growing. An Optimal Interpolation (OI) method is used to interpolate on a regular grid the hourly averaged values of temperature, relative humidity, wind vector, atmospheric pressure, and hourly cumulated precipitation. For all variables, except precipitation, the background (i.e. first guess) information is obtained by detrending the observations using the geographical parameters. For precipitation, the M. Lema radar-derived best estimate of precipitation rate at the ground is used. The characteristics of the OI schemes are shown in several test cases using data from ARPA Lombardia's mesoscale meteorological network. Finally, a quantitative diagnostics for temperature and relative humidity is carried out by using Cross Validation (CV) scores computed with large sets of data.