Convective Meteorology (Mesoscale Dynamics)

Impact of Initializing Different State Variables through Direct Insertion or Variations of Cloud Analysis on the Prediction of Convective Storm: An OSSE Study

Chong-Chi Tong
OU School of Meteorology / Center for Analysis and Prediction of Storms / Advanced Radar Research Center

17 October 2014, 2:00 PM

National Weather Center, Room 5600
120 David L. Boren Blvd.
University of Oklahoma
Norman, OK

Radar data are known to be the most important source of observations for initializing convective storms. The complex cloud analysis method used in combination with the 3DVAR has been shown to be quite effective for assimilating reflectivity data and for improving short range precipitation forecasts. Given its low computational cost, it has been used in realtime storm-scale forecasts.
Due to its semi-empirical nature and assumptions involved, the cloud analysis does not necessarily produce analysis of model state variables that are consistent with the prediction model, and often these states undergo rapid adjustments during the initial stage of forecast. Information contained in these state variables may or may not be retained well by the model after through the adjustment. It is therefore important to study the impacts and importance of the analysis of different state variables, and in particular those variables that are adjusted by the cloud analysis on the subsequent forecasts. The impacts of variations of the cloud analysis procedure should also be studied. This is best done using Observing System Simulation Experiments (OSSEs) where the truth of all state variables is known, so that the accuracy of the analyzed fields can be assessed quantitatively.
Four sets of OSSEs are conducted to explore the simulation sensitivity to four different control factors: 1) accuracy of model initial state variables, 2) model cloud microphysics, 3) cloud analysis configurations, and 4) accuracy of intercept parameter (N0) of particle size distribution (PSD) and hydrometeors distribution. Forecast results of winds (V), temperature (T), moisture (qv), and simulated radar reflectivity (Z) are examined by the root mean square error (RMSE), an index of dissimilarity between the truth simulation and experiments, within the cloudy region. Result shows that among various model state variables, accurate initial qv alone (without hydrometeor information) is able to provide long lasting (4 hours at least) good prediction of V, qv, and Z, while hydrometeors in IC are found to be crucial in creating better T forecast. Besides the initial qv accuracy, the utilization of a proper microphysics scheme with less model error is highly demanded for the benefit of initial hydrometeor on Z prediction to last longer. In the case of our cloud analysis configuration experiments, it turns out the qv adjustment in current cloud analysis procedure causes overestimated qv analysis that shortens the benefit of hydrometeor analysis on Z prediction and as well limits the forecast performance of other variables, implying the current used adjustment strategy simply based on reflectivity observation is not adequate and requires further modifications. Under the scenario of sufficient initial qv accuracy, the forecast performance difference among cloud analysis experiments with different settings of N0 and hydrometeor distribution is relatively minor yet systematic. The application of variable (predicted) N0 is found to result in giving better forecast of T, Z, and certain polarimetric variables such like ZDR and ρhv. On the other hand, a realistic hydrometeor distribution allowing coexistence of different species is required in creating more accurate ρhv analysis.
Upon the findings of current works, our future efforts will be aimed on retrieving diagnostic N0 and realistic hydrometeor distribution based on simulated radar observations and model background information. Its subsequent impact on forecast performance will be examined and verified by applications to both simulated and real cases.

For accommodations based on disability, or more details, please call 325-6561. All visitors without NOAA or University of Oklahoma identification must register at the registration desk on arrival. Visitor parking is available for all University visitors. However, faculty/staff/students must have a current multi-purpose parking permit. Additional parking is available at the Lloyd Noble Center (LNC) for those individuals who do not have a parking permit. You do not need a permit to park in one of 1,200 spaces reserved for CART bus riders, although you must ride the CART shuttle to park in the reserved area. This area is on the north central side of the Lloyd Noble Center. Elsewhere at the LNC, permits are required.

Convective Meteorology (Mesoscale Dynamics) Seminar Series website