Shortwave cloud feedback derived from a 32 year record of satellite UV reflectivity

Tuesday, 16 December 2014
Clark j Weaver, Earth System Science Interdisciplinary Center, COLLEGE PARK, MD, United States, Jay R Herman, University of Maryland JCET, Greenbelt, MD, United States and Gordon J Labow, NASA Goddard SFC, Greenbelt, MD, United States
A thirty-two year record (1980-2011) of Shortwave TOA radiative cloud forcing (Rcloud) is derived from UV reflectivity data constructed using measured upwelling radiances from Nimbus-7 SBUV and seven NOAA SBUV/2 instruments. The record of Rcloud derived from CERES is used to scale the UV reflectivity to TOA cloud forcing. The estimated linear trend for the shortwave TOA radiative forcing due to clouds from 60S to 60N is +1.1 Watts m-2 over the 32 year period.

Shortwave (SW) cloud feedback parameters are estimated using anomalies of local observed surface temperatures (Ts.) from NASA GISS. Zonal mean values of our cloud feedback parameter are similar to those derived from other data sets such as CERES, FD-ISCCP and MERRA. Over land our SW cloud feedback parameter is positive and SW Rcloud is highly correlated with Ts . However, over ocean the sign of the feedback parameters are controlled by strong negative values over equatorial latitudes balanced by positive values over subtropical and midlatitude oceans. Over ocean, the SW Rcloud response to Ts perturbations is about one month, but over land the cloud response time is less than one month.

SW cloud feedbacks estimated from all (n=384) monthly anomalies over the 32 year period are driven by ENSO variability as well as long-term warming from CO2. In an attempt to isolate the cloud feedback due to long-term global warming, we re-derive maps of the feedback but only use those months where the absolute value of the ENSO index is less than 0.5 (n=116). We compare maps of these results with SW cloud feedback estimates from CMIP5 multi-model runs that simulate an abrupt 4x increase in CO2 concentrations. Global fields from these CMIP5 model runs have many of same spatial features seen in our estimates of SW cloud feedback from long-term warming. Both show negative feedback over equatorial oceans and positive feedback over land masses.