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

C Lussana,U Pellegrini,F Uboldi,M R Salvati

doi:10.5194/asr-3-105-2009

Abstract

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.

Highlights

This document describes a spatial interpolation scheme based on Optimal Interpolation (OI; Gandin, 1963) applied to the data collected by a mesoscale meteorological network
OI is a method widely used in meteorology and climatology and it is based on a linear combination of the observed values with a background field weighting their respective uncertainties
For a review of OI and its comparison with other data interpolation techniques see for example Daley (1991) and Kalnay (2003)

Summary

Introduction

This document describes a spatial interpolation scheme based on Optimal Interpolation (OI; Gandin, 1963) applied to the data collected by a mesoscale meteorological network. Forecasters require easy access to several different variables to gain better insight into meteorological phenomena: for this reason the OI is applied to hourly averaged observations of temperature, wind, relative humidity, pressure, and to the hourly cumulated precipitation. 20 500 0.5 20 500 0.5 relative humidity main spatial trends present in the fields 3-D isotropic (dew-point temperature) and detected by the observations (*). Wind vector (w) main spatial trends present in the fields and detected by the observations. · 0.5 if |w|

Test cases

Fog in the Po Plain

Convective precipitation

Precipitation cases

Stratiform precipitation