Variational Assimilation of TMI Rain Type and Precipitation Retrievals into Global Numerical Weather Prediction

Abstract

Improvement in forecast accuracy is expected to result from the incorporation of rain type and precipitation retrieved from the TRMM Microwave Imager (TMI) into numerical weather prediction (NWP) models, especially in the tropics, where latent heating is the main energy source of atmospheric motion.A One-dimensional Variational Method (1DVAR) was developed in the present study for the assimilation of convective and stratiform rain flags, and precipitation retrieved from TMI into a Japan Meteorological Agency (JMA) global NWP system. In order to simplify the observational operators of the 1DVAR, it was assumed that the rain flags (precipitation) were functions of total water content (divergence) alone. Using these operators, the 1DVAR sought the optimal total water content (divergence), by minimizing a cost function that consists of a background total water content (divergence) term, and observation terms for the retrieved rain flags (precipitation).The impact of the TMI rain flags and precipitation assimilation on the NWP forecasts was examined in a case study of meso-scale precipitation systems over the Indian Ocean around 06 UTC 16 July 1998. The 1DVAR made deep (mid-tropospheric) moist layers with strong (moderate) updrafts in the TMIretrieved heavy rain areas (stratiform rain areas). This assimilation enabled the NWP to spin-up heavy precipitation in correspondence with the TMI retrievals, and to maintain the heavy rain areas for 24 hours. The creation of deep moist layers was essential to maintain these heavy rain areas. The assimilation also induced model stratiform precipitation around the TMI-retrieved stratiform rain areas, although the model rain areas lasted for the first several hours only.The impacts of the assimilation were also examined by including the 1DVAR into the forecastanalysis cycle of the global NWP system for July 1998. The results show that the assimilation improved large-scale patterns of precipitation, convective and stratiform rain areas, and precipitable water content over tropical oceans. It was also found that the assimilation changed double-peak profiles of diabatic Corresponding author: Kazumasa Aonashi, Mailing address: Meteorological Research Institute, 1-1 Nagamine, Tsukuba, Ibaraki, 305-0052, Japan. E-mail: Aonashi@mri-jma.go.jp ( 2004, Meteorological Society of Japan heating around inter-tropical convergence zones, into more realistic mid-tropospheric peak profiles. This change was mainly ascribable to the enhancement of tall convection and suppression of low-level convection by the 1DVAR. The Hadley circulation was strengthened (weakened) in the upper to mid-(lower-) troposphere in accordance with the changes in diabatic heating.

The Global Precipitation Measurement Mission

“…2) Enhanced numerical weather prediction skills. Assimilation of precipitation information into global and regional forecast systems has been shown to improve atmospheric analyses and short-range forecasts in a variety of situations (Zupanski et al 2002;Marécal and Mahfouf 2003;Hou et al 2004;Aonashi et al 2004). Rain-affected microwave radiances and precipitation retrievals are currently being used at NWP centers to improve operational forecasts (Bauer et al 2006).…”

Assimilation of Precipitation Information Using Column Model Physics as a Weak Constraint

“…In recent years, significant progress has been made in using these observations in data assimilation to improve atmospheric analyses and forecasts. Numerical weather prediction centers such as the NCEP, JMA, and ECMWF have begun using precipitation data or rain-affected microwave brightness temperatures in operational forecasts (Treadon et al 2002;Aonashi et al 2004;Marecal and Mahfouf 2003;Bauer et al 2006). Currently, precipitation information (either retrievals or rain-affected radiances) is assimilated in NWP systems much the same way as any other data to optimize the initial state for a better forecast.…”