Using satellite probabilistic estimates to assess modeled relative humidity : application to a NWP model

<p>Assessing model forecasts using remote sensing data is often and generally done by confronting past simulations to observations. We developed a novel probabilistic comparison method that evaluates tropical atmospheric relative humidity profiles simulated by the global numerical model for weather forecasting ARPEGE (M&#233;t&#233;o France) using probability density functions of finer scale satellite observations as reference.</p> <p>The global relative humidity field is simulated by ARPEGE every 6 hours on a 0.25 degree grid over 18 vertical levels ranging from 100 hPa to 950 hPa. The reference relative humidities are retrieved from brightness temperatures measured by SAPHIR, the passive microwave sounder onboard satellite Megha-Tropiques. SAPHIR has a footprint resolution ranging from 10 km at nadir to 23 km at the edge of the swath, with a vertical resolution of 6 vertical pressure layers (also from 100 hPa to 950 hPa).&#160; Due to the particular orbit of the satellite, each point of the Tropical belt is observed multiple times per day.&#160;</p> <p>Footprint scale RH probability density functions are aggregated (convoluted) over the spatial and temporal scale of comparison to match the model resolution and summarize the patterns over a significant period. This method allows to use more sub-grid information by considering the finer-scale distributions as a whole. Thisprobabilistic approach avoids the classical determinist simplification consisting of working with a simple &#8221;best&#8221; estimate. The resulting assessment is more contrasted while better adapted to the characterization of specific situations on a case-by-case study. It provides a significant added-value to the classical deterministic comparisons by accounting for additional information in the evaluation of the simulated field, especially for model simulations that are close to the traditional mean.</p> <p>Comparison results will be shown over the April-May-June 2018 period for two configurations of the ARPEGE model (two parametrization schemes for convection). The probabilistic comparison is discussed with respect to a classical deterministic comparison of RH values.</p>