85th AMS Annual Meeting

Monday, 10 January 2005: 2:15 PM
Assimilation of high-resolution DIAL water vapor data into the MM5 4DVAR system - Experiments and Validation
Hans-Stefan Bauer, Hohenheim University, 70599 Stuttgart, Germany; and V. Wulfmeyer, M. Grzeschik, F. Vandenberghe, and E. V. Browell
Poster PDF (2.2 MB)
The distribution and intensity of precipitation is the most important parameter in today’s weather forecasts. At the same time its prediction is very difficult since the development of precipitation occurs at the end of a long chain of processes which are only crudely represented even in high-resolution mesoscale numerical weather prediction (NWP) models.

Another reason is that the observation of the 4-dimensional distribution of water vapor is not yet accurate enough. The main source for water vapor profiles in the atmosphere are observations of radiosondes associated with different problems as coarse horizontal resolution, horizontal drift, or measurement inaccuracies above 300 hPa. The assimilation of this water vapor field introduces large errors into the initialization of numerical weather forecasts.

Lidar systems are admittedly qualified for closing existing gaps in the observation network, since they are able to observe parameters like temperature, wind and water vapor with a high temporal and spatial resolution. So far the influence of such systems on NWP, especially on the mesoscale and on QPF is not adequately investigated.

Within this project the influence of high resolution DIAL observations of the NASA LASE system on short-range forecasts of the mesoscale NWP model MM5 should be investigated with assimilation experiments for a convective situation. The observations were collected during the International Water Vapor Project (IHOP_2002). LASE provides water vapor profiles of the troposphere with a high spatial and temporal resolution as well as accuracy. To assimilate the observations into the model a continuous and physically consistent assimilation scheme is necessary. Here the new 4DVAR system, based on version 3.4 of MM5 is used.

The development of the necessary observation operator and the methodology to assimilate Lidar water vapor data is introduced in more detail in the talk of Wulfmeyer et al. First experiments in which only LASE data was continuously assimilated showed clear positive effects on the development of convection in the forecast. As observed convection occurs further to the south compared with the control simulation and this positive signal is clearly seen in the forecast hours after the end of the assimilation window.

In this work, validations of the water vapor and precipitation fields against independent observations are performed. Also results of different assimilation experiments for the selected case study are presented considering additional observing systems supported by the 4DVAR system. This enables the investigation of the impact of different observing systems and of the value of DIAL water vapor data. First results are presented at the conference.

Supplementary URL: