Files in this item



application/pdfFengxue_Qiao.pdf (5MB)
(no description provided)PDF


Title:CWRF summer precipitation prediction over the United States land and coastal oceans: effects of ensemble cumulus parameterization closures
Author(s):Qiao, Fengxue
Director of Research:Liang, Xin-Zhong
Doctoral Committee Chair(s):Liang, Xin-Zhong
Doctoral Committee Member(s):Wuebbles, Donald J.; McFarquhar, Greg M.; Kumar, Praveen
Department / Program:Atmospheric Sciences
Discipline:Atmospheric Sciences
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):cumulus parameterization
regional climate model
precipitation prediction
Abstract:A long-standing problem in climate modeling is the accurate prediction of precipitation with respect to three key characteristics: geographic variation of total amount, frequency and intensity, and the diurnal cycle. In particular, coarse-resolution climate models have low predictive skills at regional scales, while the higher-resolution regional climate/weather models show improved downscaling skill but still exhibit biases and are highly sensitive to cumulus parameterization (CUP) especially during the summer. Most current models also tend to predict rainfall too early in the daytime and too frequently at the light intensity over both land and oceans. These problems have been identified with model deficiencies in CUP. As the core problem of CUP, cumulus closure assumptions fundamentally determine the location, frequency and intensity of convective rainfall. Numerous closures have been proposed, but there is no consensus with respect to their relative performance. This study uses the CWRF model which incorporates an Ensemble Cumulus Parameterization (ECP) scheme to evaluate the performance of different widely-used closures for summer precipitation prediction regarding the above three key features. The ECP includes five major groups of closure assumptions with 16 different algorithms: the Arakawa-Schubert quasi-equilibrium (AS), the vertical velocity (W), the moisture convergence (MC), the total instability adjustment (KF), and the instability tendency (TD). Extensive experiments are conducted by implementing these closures separately over the continental U.S. and adjacent coastal oceans. Results show that cumulus closures significantly affect U.S. precipitation patterns, heavy rainfall occurrence, and the diurnal cycle, with strong regional dependence differing between land and coastal oceans. Over the U.S. coastal oceans, two closure algorithms using the average vertical velocity at the cloud base (W_2) and moisture convergence (MC_3) complementarily reproduce the summer precipitation patterns and amount, and both skillfully capture the frequency of heavy rainfall events. However, the instability tendency closures are superior in capturing the diurnal phase but with much larger amount deficits. This suggests that cloud base vertical velocity and moisture convergence primarily determine seasonal mean and daily precipitation variability, but the instability tendency plays a critical role in regulating precipitation sub-daily variation. Over the continental U.S., the MC closure most realistically reproduces Central U.S. summer rainfall amount, daily precipitation variation and frequency distribution, but produces wet biases over the North American Monsoon (NAM) region and Southeast U.S. which can be significantly reduced by using the W closure. Further skill enhancement can be made using an optimized ensemble of the MC and W closures. The TD and KF closures show advantages in capturing the diurnal signals over the Central U.S. and NAM, respectively. This reasonably explains the systematic behaviors of several major CUP schemes. This research further compares the performance of CWRF using the ECP with other 11 CUP schemes in predicting the Central U.S. summer floods. The ECP scheme with the MC and W_2 closures separately over the land and oceans shows advantages over other schemes in simulating the Central U.S. flood amount, daily rainfall frequency and intensity. The Grell scheme shows superiority in reproducing Central U.S. nocturnal rainfall maxima, but other schemes generally fail. This advantage of the Grell scheme is primarily due to the instability tendency closure assumption. Future studies will attempt to incorporate this instability tendency assumption as a trigger function in the ECP scheme to improve the Central U.S. rainfall diurnal simulations.
Issue Date:2013-05-24
Rights Information:Copyright 2013 Fengxue Qiao
Date Available in IDEALS:2013-05-24
Date Deposited:2013-05

This item appears in the following Collection(s)

Item Statistics