Skip to Main Navigation. Each navigation link will open a list of sub navigation links.

Skip to Main Content

Apply Now to the University of Wyoming apply now
Visit Campus
Download UW Viewbook
Give to UW

PhD Defense

Department of Atmospheric Science

Tue., Mar. 5, 3:10 pm, EN6085

Evaluation and improvement of the MODIS liquid water path retrievals using A-train satellite and ground-based remote sensing measurements and radiative transfer modelling

Sujan Khanal

University of Wyoming

Abstract

Cloud liquid water path (LWP) strongly influences both radiative and hydrological properties of clouds and satellite-based LWP retrievals provide essential information for characterizing the state of the climate system and for evaluating climate models. However, current satellite LWP retrievals can contain significant biases that are too large to properly constrain inter-model variation. In this study, three years of combined A-train remote sensing measurements from MODIS, AMSR-E, CloudSat and CALIPSO, combined with ground-based measurements and three dimensional radiative-transfer modelling are used to examine uncertainties in MODIS LWP retrievals. MODIS retrievals, based on a bi‑spectral reflectance technique, are unique in that they can provide global daytime observations of both liquid and ice cloud properties at a spatial resolution of 1 km every one to two days. The focus of this study is to improve the retrieval of LWP from MODIS measurements for both mixed phase clouds and clouds with high solar zenith angle, both of which disproportionately affect the MODIS retrievals at high latitudes.

Current MODIS retrievals are based on either liquid or ice phase assumption throughout the cloud vertical layer. This study improves the operational approach in liquid-topped mixed phase clouds by adding complementary information from active sensors onboard the A-train. They allow proper identification of cloud phase and also provide independent measurements of ice clouds properties. Multi-sensor LWP retrievals in liquid topped mixed phase clouds show that inadequate phase classification contribute to the LWP bias that is related to the IWP on average and reaches close to 15% at IWP of 150 g/m2 and can reach 40% or higher when IWP is greater than 400 g/m2.

Moreover, A-train and ground-based measurements and radiative transfer modeling are used to quantify the as-of-yet unresolved high bias in MODIS LWP that depends on solar zenith angle. This study confirms that cloud top height inhomogeneity is one of the main factors that contribute to this bias due to 3-dimensional radiative effects. A new approach is introduced to reduce this bias. It utilizes two parameters: the solar zenith angle and cloud heterogeneity index, both of which are provided in the cloud property dataset for each MODIS pixel. Comparisons with collocated satellite and ground-based microwave measurements in carefully screened clouds shows that this method can effectively remove this large first-order MODIS LWP bias. Applying this to MODIS pixel level data collocated along the CloudSat footprint suggests annually averaged MODIS LWP overestimation gradually increases with latitude and can reach over 50 g/m2 at high latitudes. This overestimation becomes even more severe for seasonal or monthly averages and can exceed 100 g/m2 during winter. An improved MODIS LWP dataset will provide an observational constraint on the temperature-dependent cloud phase partitioning in models and can help improve cloud feedback uncertainty in mixed phase clouds, which are prevalent in polar regions where observational data are limited and contain large uncertainties.

 

 

 

 

Contact Us

University of Wyoming,

Atmospheric Science,

EN 6034

Dept. 3038

1000 E. University Ave.

Laramie, WY 82071

Phone: (307)766-3245

Email: geerts@uwyo.edu

1000 E. University Ave. Laramie, WY 82071
UW Operators (307) 766-1121 | Contact Us | Download Adobe Reader

Accreditation | Virtual Tour | Emergency Preparedness | Employment at UW | Privacy Policy | Harassment & Discrimination | Accessibility Accessibility information icon