Precipitation and Planetary Boundary Layer Height Analysis Using Surface-Based, Airborne, and Spaceborne Measurements

Abstract

Atmospheric science has grown increasingly important, with fields such as numerical weather forecasts directly impacting people’s daily lives. However, limitations on the quality of observed precipitation and planetary boundary layer (PBLH) produce challenges in fields such as weather forecasting, extreme precipitation, and pollution transportation. Furthermore, the limitation of cloud microphysics in climate models continues to be an important topic. Despite gaining increased attention in recent years, observational challenges and limited understanding of cloud microphysics remain significant over the ocean, an area with scarce surface-based and airborne data. Given the above challenges, this dissertation aims to evaluate observational data and methods to estimate rainfall and PBLH in regions such as over the ocean with limited data. It also uses high-quality datasets to study their climatology, providing valuable insights to improve these datasets. Another goal is to use observation-based parameterization of cloud drop size distribution to explore its impact on radiation and cloud microphysics in climate models. This dissertation includes four methods: data analysis, algorithm evaluations, algorithm improvement, and model sensitivity tests. Using the above techniques, we tackled, evaluated, and improved the ways to understand observational datasets, improve current retrieval algorithms, and quantify changes in model outputs using new parameterizations informed from data analysis. To address the above objectives, this dissertation presents the following findings: 1. Precipitation Analysis: Gauge-corrected radar data over the U.S. showed that coastal land receives more rainfall than coastal ocean. An evaluation of three satellite precipitation products (IMERG, PERSIANN, and CMOPRH) using gauge-corrected radar data showed that IMERG performs best over coastal land, and CMORPH performs best over coastal oceans. 2. PBLH estimation: A new algorithm was developed for estimating PBLH from dropsondes’ thermodynamic profile over the northwest Atlantic. We also evaluated the mixed layer height product from the airborne High Spectral Resolution Lidar-Generation 2 (MLH-HSRL) and found that it agrees well with that of dropsondes. Furthermore, we improved the MLH-HSRL to estimate PBLH better. For the other study, we evaluated four methods estimating PBLH from dropsonde profiles: We first identified the best one for each of the available four methods, and then we identified the parcel method and the gradient Richardson number as the two best methods estimating PBLH with the parcel method performing slightly better. 3. Model sensitivity test: Sensitivity test demonstrates that the new parameterization of cloud droplet size distribution from Siu et al. (2025) data analysis improves summer Arctic shortwave cloud forcing in the Community Earth System Model simulation by 5 W/m2. This dissertation provides new angles for precipitation evaluation and contributes to the broader use and insights of airborne lidar and dropsondes to estimate PBLH. The findings can be implemented in future field campaigns and global satellite missions. Additionally, insights from the climate model sensitivity test highlight the importance of integrating observational data for model improvement of cloud microphysics on radiative feedback. Overall, this dissertation takes an approach from remote sensing to model parameterization, forming a loop from observation and model improvement. Hence, through data analysis, algorithm evaluation and improvement, and model sensitivity tests, this dissertation offers a meaningful basis for improving the understanding of atmospheric science.