Date of Award

1-1-2021

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Atmospheric and Environmental Sciences

Content Description

1 online resource (xiv, 129 pages) : color illustrations, color maps.

Dissertation/Thesis Chair

Qilong M Min

Committee Members

David Fitzjarrald, Brian Rose, Liming Zhou

Keywords

Atmospheric models, Weather forecasting, Precipitation forecasting, Cloud forecasting, Boundary layer (Meteorology), Earth temperature

Subject Categories

Atmospheric Sciences

Abstract

The land-atmosphere coupling system is important for the simulation of key quantities like surface temperature, precipitation, and radiative energy. Over the complex terrain of New York State, the land-atmosphere coupling process is quite complex and misrepresenting the coupling processes could lead to strong biases. Evaluating the weather forecasting models is vital for enhancing understanding of physical and processes and further improving the model forecasting. A comprehensive observation network, the New York State Mesonet (NYSM) provides a great opportunity to investigate how the land atmosphere coupling process are simulated over complex terrain region. This research includes three components. In first part, HRRR (High-Resolution Rapid Refresh) model surface meteorology, soil hydrological process, surface energy partition and cloud condition are evaluated for an entire year. The model has been evaluated from both temporal (seasonal and diurnal) and spatial perspectives. The results demonstrate that the HRRR model surface thermodynamic biases are seasonally dependent, presenting a systematic warm and dry bias during the warm season and extreme cold biases in the coldest days. The warm season biases are controlled by the land use types and the presence of optically thick clouds. A hydrological bias underestimating of spring snowmelt infiltration water partly explains the subsequent summer warm and dry biases over farmland. The diurnal temperature bias is partly controlled by the modelled cloud fraction. The daytime temperature bias is largest and nocturnal temperature bias is smallest when the cloud fraction is high in the model. In the second part, using the Weather Research and Forecasting (WRF) model, ten sensitivity tests were run to test the impact of 1) the model resolution, 2) the model physical parameterizations of land surface and boundary layer, and 3) the model parameterization of sub-grid scale process on surface meteorology and boundary layer cloud development. The results show that over the complex terrain, the high-resolution simulations (1-km × 60-level) generally performs better with more details compared to a low-resolution simulation (3-km × 50-level) in both surface meteorology and cloud fields. The surface meteorology in the model is more sensitive to Land Surface Models (LSMs) than Planetary Boundary Layer (PBL) schemes. NoahMP land surface model exhibits warmer and drier biases compared to Rapid Update Cycle (RUC). The PBL schemes control the coupling strength of land surface and atmosphere. With stronger coupling strength, non-local schemes tend to produce more cloud coverage. However, by considering the radiation effect of Subgrid Scale (SGS) clouds, Mellor-Yamada-Nakanishi-Niino (MYNN) predict highest cloud coverage and lowest surface solar radiation bias. In the third part, the surface meteorology, fluxes, and cloud development are assessed during a dry and a wet year. The coupling between surface and atmosphere is diagnosed using a mixing diagram. The results show that the warm season dry and warm biases systematically increased during a drought year, especially for the forest sites. The model simulation of latent heat flux is too sensitive to the drought condition and is largely underestimated in the model in drought year. During days with boundary layer clouds, the mixing diagram shows that the moist static energy grows much slower in the model compared to observation. It is possible that this biases partly attributed to the underestimation of cloud optical depth due to not enough energy for the cloud development.

Download

Share

COinS