Estimations of column water amounts over Arctic regions using ground-based thermal microwave and infrared measurements

Over the Arctic environments, ARM cloud liquid water path (LWP) measurements using ground-based microwave radiometer (MWR) will be critical for understanding the properties of clouds in the Arctic and for verifying satellite retrievals of cloud properties. These estimates from the MWR data can have large errors due to both cold water temperature (Tw) and small column water vapor (CWV) amount. Based on comparisons with airborne cloud microphysical measurements during the SHEBA/FIRE/ARM experiment, the standard ARM LWP estimates are larger by about a factor of two. The microwave absorption coefficients for supercooled water are significantly larger than those for warm (> 0 deg. C) water. For the small CWV cases, the contribution of microwave gas continuum absorption to the total gas microwave absorption is relatively greater than that in humid conditions. Different microwave gas absorption models result large differences in the gas continuum absorption. Thus, the selection of the atmospheric gas and water absorption models and determination of cloud water temperature are critical for LWP retrievals using MWR in polar regions. This study uses the upper-looking Infrared (IR) Thermometer (IRT) and MWR at the SHEBA/ARM site to estimate LWP, CWV and Tw simultaneously. Current method physically accounts for atmospheric gas and cloud water absorptions at the MWR microwave wavelengths using the radiative transfer model of Lin et al. [1998]. The microwave absorption coefficients for gas and water are calculated from the models of Liebe [1989] and Ray [1972], respectively. The cloud water temperature estimation is made from a parameterization based on physical calculations of IR radiative transfer using correlated K distribution [Kratz et al. 1998]. This method results a about 25 ~ 45 % reduction in LWP values comparing with the standard ARM estimates. If extreme large values (LWP > 0.5 mm) which are likely associated with precipitation are excluded, current results (~0.033 mm) are very close to the aircraft measurements (~0.034 mm). The correlation between the two is also very good, i.e., its coefficient is statistically above the 95% significant level. Compared with the original SHEBA/ARM CWV values, the changes in CWV are small (< 5% relatively). This method provides great potential for ground-based cloud and atmosphere remote sensing.