Li, Yue (2012-02). Investigation of the Dynamical, Macrophysical and Radiative Properties of High Clouds Combining Satellite Observations and Climate Model Simulations. Doctoral Dissertation.
Thesis
This dissertation investigates three topics concerning high clouds: 1) convectively coupled equatorial wave (CCEW) signals derived from cloud top temperature (CTT) and cirrus optical thickness retrieved from satellite observations; 2) investigation of the physical mechanism governing the fixed anvil temperature (FAT) hypothesis and test of FAT hypothesis with CTT measurements; and 3) the intercomparison of cloud fraction and radiative effects between satellite-based observations and reanalysis product and simulations from general circulation models (GCMs). A wealth of information on CCEWs is derived from Aqua/MODIS cloud-top properties. We apply space-time spectral analysis on more than 6 years of CTT and isolate various modes of CCEWs including Kelvin, n = 1 equatorial Rossby, mixed Rossby-gravity, n = 0 eastward inertio-gravity waves, and the Madden-Julian oscillation. The successful application of the same method to cirrus cloud optical thickness confirms robust convective signals at upper troposphere. Consistent with the physical governing mechanism of the FAT hypothesis, the peak clear-sky diabatic subsidence, convergence and cloud fraction are located at roughly the same level (200 hPa), which is fundamentally determined by the rapid decrease of water vapor concentration above this level. The geographical maxima of cloud fraction agree well with that of water vapor, clear-sky cooling rates, diabatic subsidence and convergence at 200 hPa. An analysis of the response of the tropical mean CTT anomaly time series to sea surface temperature indicates that a possible negative relationship is present. In addition, we suggest interpreting the FAT hypothesis, and the more recent proportionately higher anvil temperature (PHAT) hypothesis, by using the temperature at the maximum cloud detrainment level instead of the CTT. Simulations of cloud fraction and radiative properties using two versions of the NCAR CAM models indicate that an overall improvement is observed in CAM5 compared to CAM3. However, an apparent bias in CAM5 shortwave (SW) cloud radiative forcing (CRF) simulation is shown in boreal winter southern mid latitude. This bias is primarily due to the underestimation of fraction-weighted SW CRF related to both high and middle top clouds. Additionally, apparent compensation errors are observed in models.
