MODIS Cloud Detection Evaluation Using CALIOP over Polluted Eastern China

Saichun Tan,Guangyu Shi,Xiao Zhang

doi:10.3390/atmos10060333

Abstract

Haze pollution has frequently occurred in winter over Eastern China in recent years. Over Eastern China, Moderate Resolution Imaging Spectroradiometer (MODIS) cloud detection data were compared with the Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) for three years (2013–2016) for three kinds of underlying surface types (dark, bright, and water). We found that MODIS and CALIOP agree most of the time (82% on average), but discrepancies occurred at low CALIOP cloud optical thickness (COT < 0.4) and low MODIS cloud top height (CTH < 1.5 km). In spring and summer, the CALIOP cloud fraction was higher by more than 0.1 than MODIS due to MODIS’s incapability of observing clouds with a lower COT. The discrepancy increased significantly with a decrease in MODIS CTH and an increase in aerosol optical depth (AOD, about 2–4 times), and MODIS observed more clouds that were undetected by CALIOP over PM2.5 > 75 μg m−3 regions in autumn and particularly in winter, suggesting that polluted weather over Eastern China may contaminate MODIS cloud detections because MODIS will misclassify a heavy aerosol layer as cloudy under intense haze conditions. Besides aerosols, the high solar zenith angle (SZA) in winter also affects MODIS cloud detection, and the ratio of MODIS cloud pixel numbers to CALIOP cloud-free pixel numbers at a high SZA increased a great deal (about 4–21 times) relative to that at low SZA for the three surfaces. As a result of the effects of aerosol and SZA, MODIS cloud fraction was 0.08 higher than CALIOP, and MODIS CTH was more than 2 km lower than CALIOP CTH in winter. As for the cloud phases and types, the results showed that most of the discrepancies could be attributed to water clouds and low clouds (cumulus and stratocumulus), which is consistent with most of the discrepancies at low MODIS CTH.

Highlights

Clouds play an important role in the global climate and contribute large uncertainty to estimates and interpretations of the Earth’s changing energy budget [1,2]
The purpose of this study is to investigate what differences occurred between passive Moderate Resolution Imaging Spectroradiometer (MODIS) and active Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) cloud detection over Eastern China, with particular focus on the following: (1) The MODIS cloud detection evaluation over high aerosol concentration regions using
Because of the sparse spatial sampling from Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) [15,16], when calculating the regional distribution of cloud fraction, solar zenith angle (SZA) and cloud optical thickness (COT), we considered all MODIS and CALIOP pixels in each 2.5◦ × 2.5◦ box to ensure more data were included

Summary

Introduction

Clouds play an important role in the global climate and contribute large uncertainty to estimates and interpretations of the Earth’s changing energy budget [1,2]. Clouds can reflect the incoming solar radiation and have a cooling effect for the climate. Clouds absorb the longwave radiation emitted by the Earth and emit energy to space. The greenhouse effect of clouds is much larger than that from a CO2 doubling [1,3]. It continues to be a challenge to quantify the cloud and convective effect in computer models [4,5,6]. It was found that a small change in some cloud properties

Objectives

Results

Conclusion