The impact of assimilating lower‐tropospheric lidar temperature profiles into a numerical weather prediction (NWP) model was investigated. The profiles were measured with the Temperature Rotational Raman Lidar (TRRL) of the University of Hohenheim on 24 April 2013. The day showed the development of a typical daytime planetary boundary layer (PBL) with no optically thick clouds. The Weather Research and Forecasting (WRF) model was operated with 57 vertical levels covering Central Europe with 3 km horizontal resolution. Three different experiments were carried out with a rapid update cycle with hourly three‐dimensional variational data assimilation. The impact run (ALL_DA) was performed with the assimilation of conventional data and the additional assimilation of TRRL profiles between 0900 and 1800 UTC in a height range from about 500 to 3000 m above ground level with a vertical resolution of about 100 m. In CONV_DA and NO_DA, only conventional data and no data were assimilated, respectively. To consider the representativeness of the TRRL profiles, an observation error of 0.7 K was used for all heights. The assimilation was performed using the radiosonde operator. The TRRL data assimilation corrected the temperature profiles towards the lidar data. In the mean, the boundary‐layer height was improved by 60 m in ALL_DA compared to the TRRL data and the temperature gradient in the entrainment layer by 0.19 K (100 m)−1. While ALL_DA showed a root mean square error (RMSE) of 0.6 K compared to the TRRL data, the RMSE of CONV_DA was twice as large. Compared to data from radiosondes launched at the TRRL site, ALL_DA showed a significantly smaller RMSE than CONV_DA in two out of the four times radiosonde data were available. We conclude that the assimilation of TRRL data has great potential to close the critical gap of missing temperature observations in the lower troposphere.
The impact of assimilating thermodynamic profiles measured with lidars into the Weather Research and Forecasting (WRF)-Noah-Multiparameterization model system on a 2.5-km convection-permitting scale was investigated. We implemented a new forward operator for direct assimilation of the water vapor mixing ratio (WVMR). Data from two lidar systems of the University of Hohenheim were used: the water vapor differential absorption lidar (UHOH WVDIAL) and the temperature rotational Raman lidar (UHOH TRL). Six experiments were conducted with 1-hour assimilation cycles over a 10-hour period by applying a 3DVAR rapid update cycle (RUC): 1) no data assimilation 2) assimilation of conventional observations (control run), 3) lidar−temperature added, 4) lidar−moisture added with relative humidity (RH) operator, 5) same as 4) but with the WVMR operator, 6) both lidar−temperature and moisture profiles assimilated (impact run). The root-mean-square-error (RMSE) of the temperature with respect to the lidar observations was reduced from 1.1 K in the control run to 0.4 K in the lidar−temperature assimilation run. The RMSE of the WVMR with respect to the lidar observations was reduced from 0.87 g kg −1 in the control run to 0.53 g kg −1 in the lidar−moisture assimilation run with the WVMR operator, while no improvement was found with the RH operator; it was reduced further to 0.51 g kg −1 in the impact run. However, the RMSE of the temperature in the impact run did not show further improvement. Compared to independent radiosonde measurements, the temperature assimilation showed a slight improvement of 0.71 K in the RMSE to 0.63 K, while there was no conclusive improvement in the moisture impact. The correlation between the temperature and WVMR variables in the static-background error-covariance matrix affected the improvement in the analysis of both fields simultaneously. In the future, we expect better results with a flow-dependent error covariance matrix. In any case, the initial attempt to develop an exclusive thermodynamic lidar operator gave promising results for assimilating humidity observations directly into the WRF data assimilation system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.