Author:
Huang D.,Gasiewski A.,Wiscombe W.
Abstract
Abstract. Part 1 of this research reveals that many conditions of the 2003 cloud tomography experiment at the Wakasa Bay were not ideal for the tomographic retrieval purpose. For example, the aircraft flew too fast and its altitude was too high. Part 2 (this paper) then focuses on the examination of several possible improvements on the mobile cloud tomography method by means of observation system simulation experiment. We find that the incorporation of the L1 norm total variation regularization in the tomographic retrieval algorithm better reproduces small scale discontinuous structure than the widely used L2 norm Tikhonov regularization and successfully resolves sharp cloud edges. The simulation experiments reveal that a typical ground-based mobile cloud tomography setup substantially outperforms an airborne one because of its slower moving speed and greater contrast in microwave brightness between clouds and the cosmic background. The simulations show that, as expected, the retrieval error increases monotonically with radiometer noise level and the uncertainty in background brightness temperature. It is also revealed that a slower platform or a faster scanning radiometer results in more scan cycles and better overlapping between the swaths of successive scan cycles, both of which are highly favorable for cloud tomography retrieval. The last factor examined is aircraft height. It is shown that the best retrieval is obtained when the aircraft data are collected at the altitudes between 500 m to 1000 m above the cloud top. To summarize, this research demonstrates the feasibility of tomographically retrieving cloud structure using current scanning microwave radiometer technology and provides several general guidelines by which to improve future field-based studies of cloud tomography.