Assimilation of wind speed and direction observations: results from real observation experiments

Abstract

A B S T R A C TThe assimilation of wind observations in the form of speed and direction (asm_sd) by the Weather Research and Forecasting Model Data Assimilation System (WRFDA) was performed using real data and employing a series of cycling assimilation experiments for a 2-week period, as a follow-up for an idealised post hoc assimilation experiment. The satellite-derived Atmospheric Motion Vectors (AMV) and surface dataset in Meteorological Assimilation Data Ingest System (MADIS) were assimilated. This new method takes into account the observation errors of both wind speed (spd) and direction (dir), and WRFDA background quality control (BKG-QC) influences the choice of wind observations, due to data conversions between (u,v) and (spd, dir). The impacts of BKG-QC, as well as the new method, on the wind analysis were analysed separately. Because the dir observational errors produced by different platforms are not known or tuned well in WRFDA, a practical method, which uses similar assimilation weights in comparative trials, was employed to estimate the spd and dir observation errors. The asm_sd produces positive impacts on analyses and short-range forecasts of spd and dir with smaller root-mean-square errors than the u,v-based system. The bias of spd analysis decreases by 54.8%. These improvements result partly from BKG-QC screening of spd and dir observations in a direct way, but mainly from the independent impact of spd (dir) data assimilation on spd (dir) analysis, which is the primary distinction from the standard WRFDA method. The potential impacts of asm_sd on precipitation forecasts were evaluated. Results demonstrate that the asm_sd is able to indirectly improve the precipitation forecasts by improving the prediction accuracies of key wind-related factors leading to precipitation (e.g. warm moist advection and frontogenesis).

Impact of Assimilation of Satellite Retrieved Ocean Surface Winds on the Tropical Cyclone Simulations Over the North Indian Ocean

“…Gao et al. (2015) assimilated satellite‐derived Atmospheric Motion Vectors (AMV) and surface data set in Meteorological Assimilation Data Ingest System (MADIS) in WRFDA, which reduces root mean square error (RMSE) in wind speed and direction analyses. The improvement for wind speed analyses was 3.7%, and that for wind direction analyses was 6.9%.…”

Impact of SCATSat‐1 Retrieved Wind Vectors on Short‐Range WRF Model Predictions Over the South Asia Region

“…In all assimilation experiments, UV components are used for assimilation in which error in wind speed is considered as U and V components error, a default procedure for WRFDA assimilation system. Huang et al () and Gao et al () presented that assimilation of wind speed and direction are better option due to precise use of wind speed and direction error for upper air winds that may be a scope for further research for surface winds assimilation. In the study, all observation errors are assumed uncorrelated in space and time.…”

Land Surface Parameter and State Perturbations in the Global Ensemble Forecast System

“…5c,g). When assuming an observational error of about 1 m s 21 (Gao et al 2015), spread and RMSE were similar in magnitude for almost all lead times. The spread increases from the surface perturbations were very small, on the order of 0.05 m s 21 and only E ssppt and E all had a significant impact.…”