Comparing the assimilation of visible and infrared satellite observations to radar reflectivity for convective-scale numerical weather prediction

Description

Early warning of storms requires early detection and assimilation of convective-scale features including clouds. Visible satellite imagery detects deep convection earlier than radar, but is largely overlooked by research and operational centers (apart from DWD). We estimate the potential impact of assimilating cloud-affected satellite observations in the visible (0.6 µm) and near thermal infrared wavelengths (6.2 µm and 7.3 µm) relative to the impact of assimilating radar reflectivity observations. For that purpose, we performed idealized observing system simulation experiments (OSSE) using the Weather Research and Forecasting model (WRF) at 2-km grid resolution, the radiative transfer model RTTOV/MFASIS, and the Ensemble Adjustment Kalman Filter in the Data Assimilation Research Testbed (DART). The forecast impact was evaluated in two cases: isolated and scattered supercells with different assumed prior uncertainties.
The main result is that satellite observations can be nearly as beneficial as three-dimensional radar reflectivity observations, depending on the prior forecast uncertainty. We conclude that there is a high potential impact of assimilating cloud-affected satellite observations when the location of convection is uncertain and comparatively less potential when the stage of convective development is uncertain. Additionally, we are studying the effect of observation operator nonlinearity on the analysis mean and variance update to investigate undesirable systematic effects.

Period	17 Oct 2023
Event title	9th International Symposium on Data Assimilation
Event type	Conference
Location	Bologna, ItalyShow on map
Degree of Recognition	International