Potential of microwave observations for the evaluation of rainfall and convection in a regional climate model in the frame of HyMeX and MED-CORDEX

Abstract

This study evaluates the potential of spaceborne passive microwave observations for assessing decadal simulations of precipitation from a regional climate model through a model-to-satellite approach. A simulation from the Weather and Research Forecasting (WRF) model is evaluated against 2002-2012 observations from the Advanced Microwave Sounding Unit (AMSU-B) and the Microwave Humidity Sounder (MHS) over the Mediterranean region using the radiative transfer code RTTOV (Radiative Transfer for Tiros Operational Vertical Sounder). It is first shown that simulated and observed brightness temperatures are consistently correlated for both water vapour and window channels. Yet, although the average simulated and observed brightness temperatures are similar, the range of bright- ness temperatures is larger in the observations. The difference is presumably due to the too low content of frozen particles in the simulation. To assess this hypothesis, density and altitude of simulated frozen hydrometeors are compared with observations from an airborne cloud radar. Results show that simulated frozen hydrometeors are found at lower median altitude than observed frozen hydrometeors, with an average content at least 5 times inferior. Spatial distributions of observed and simulated precipitation match reasonably well. However, when using simulated brightness temperatures to diagnose rainfall, the simulation performs very poorly. These results highlight the need of providing more realistic frozen hydrometeors content, which will increase the interest of using passive microwave observations for the long-term evaluation of regional models. In particular, significant improvements are expected from the archiving of convective fluxes of precipitating hydrometeors in future regional model simulation programs.

Editorial for the Med-CORDEX special issue

“…As an illustration, Chakroun et al (2016) and Rysman et al (2016) test new satellite products for regional model evaluation using respectively lidar-based and microwavebased spatial products to evaluate the vertical structure of cloud and rainfall over the sea. Adloff et al (2017) use for the first time the new ESA-CCI product for the Mediterranean Sea level.…”

“…More specific evaluations of the extreme precipitations are conducted by Drobinski et al (2016b) focusing on the scaling with local temperature on stations, by Khodayar et al (2016) focusing on one of the HyMeX case study in Fall 2012 and by Rysman et al (2016) using satellite-derived extreme precipitation indices over sea.…”

All‐sky satellite data assimilation at operational weather forecasting centres

“…For example, one technical reason why NCEP started all‐sky assimilation in non‐precipitating cloudy areas was because the precipitation forecast fields were not easily available in their system. In the context of climate monitoring, Rysman et al () have illustrated the very large biases that result from simulating precipitation‐affected microwave brightness temperatures without access to the vertical profiles of convective precipitation. Investigators wishing to simulate all‐sky radiances offline using archived ECMWF forecast fields cannot do this, because the convective fluxes are not archived (another issue is that not every model time step is stored).…”

Evaluation of microwave radiances of GPM/GMI for the all-sky assimilation in RTTOV framework

“…A study by Wu et al, (2015) found that frozen hydrometeors are underestimated in WRF simulations by all convective parametrization schemes over central and eastern pacific region. Rysman et al, (2016) estimate the underestimation in WRF simulations by a factor of 5 using airborne radar in Mediterranean region.…”

“…Though the overall pattern and location of convective clouds near the eye of cyclone matched closely with the observations, Tb inside the core can be found to vary with hydrometeor shapes and estimates. This may be attributed to deficiency of frozen hydrometeors at sub-grid scale in Kain-Fritsch convection scheme (Rysman et al, 2016). A study by Wu et al, (2015) found that frozen hydrometeors are underestimated in WRF simulations by all convective parametrization schemes over central and eastern pacific region.…”