Premium
On the response of MODIS cloud coverage to global mean surface air temperature
Author(s) -
Yue Qing,
Kahn Brian H.,
Fetzer Eric J.,
Wong Sun,
Frey Richard,
Meyer Kerry G.
Publication year - 2017
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1002/2016jd025174
Subject(s) - cloud cover , moderate resolution imaging spectroradiometer , cloud computing , cloud top , environmental science , cloud fraction , remote sensing , spectroradiometer , anomaly (physics) , meteorology , satellite , atmospheric sciences , climatology , geography , geology , computer science , reflectivity , physics , optics , condensed matter physics , operating system , engineering , aerospace engineering
The global surface temperature change (Δ T s ) mediated cloud cover response is directly related to cloud‐climate feedback. Using satellite remote sensing data to relate cloud and climate requires a well‐calibrated, stable, and consistent long‐term cloud data record. The Collection 5.1 (C5) Moderate Resolution Imaging Spectroradiometer (MODIS) cloud observations have been widely used for this purpose. However, the MODIS data quality varies greatly with the surface type, spectral region, cloud type, and time periods of study, which calls for additional caution when applying such data to studies on cloud cover temporal trends and variability. Using 15 years of cloud observations made by Terra and Aqua MODIS, we analyze the Δ T s ‐mediated cloud cover response for different cloud types by linearly regressing the monthly anomaly of cloud cover (Δ C ) with the monthly anomaly of global T s . The Collection 6 (C6) Aqua data exhibit a similar cloud response to the long‐term counterpart simulated by advanced climate models. A robust increase in altitude with increasing Δ T s is found for high clouds, while a robust decrease of Δ C is noticed for optically thick low clouds. The large differences between C5 and C6 results are from improvements in calibration and cloud retrieval algorithms. The large positive cloud cover responses with data after 2010 and the strong sensitivity to time period obtained from the Terra (C5 and C6) data are likely due to calibration drift that has not been corrected, suggesting that the previous estimate of the short‐term cloud cover response from the these data should be revisited.