Statistical retrieval of thin liquid cloud microphysical properties using ground‐based infrared and microwave observations
详细信息 查看全文
- 作者：Tobias Marke ; Kerstin Ebell ; Ulrich Löhnert ; David D. Turner
- 刊名：Journal of Geophysical Research: Atmospheres
- 出版年：2016
- 出版时间：27 December 2016
- 年：2016
- 卷：121
- 期：24
- 页码：14,558-14,573
- 全文大小：3.4MB
- ISSN：2169-8996
文摘
In this article, liquid water cloud microphysical properties are retrieved by a combination of microwave and infrared ground-based observations. Clouds containing liquid water are frequently occurring in most climate regimes and play a significant role in terms of interaction with radiation. Small perturbations in the amount of liquid water contained in the cloud can cause large variations in the radiative fluxes. This effect is enhanced for thin clouds (liquid water path, LWP <100 g/m2), which makes accurate retrieval information of the cloud properties crucial. Due to large relative errors in retrieving low LWP values from observations in the microwave domain and a high sensitivity for infrared methods when the LWP is low, a synergistic retrieval based on a neural network approach is built to estimate both LWP and cloud effective radius (reff). These statistical retrievals can be applied without high computational demand but imply constraints like prior information on cloud phase and cloud layering. The neural network retrievals are able to retrieve LWP and reff for thin clouds with a mean relative error of 9% and 17%, respectively. This is demonstrated using synthetic observations of a microwave radiometer (MWR) and a spectrally highly resolved infrared interferometer. The accuracy and robustness of the synergistic retrievals is confirmed by a low bias in a radiative closure study for the downwelling shortwave flux, even for marginally invalid scenes. Also, broadband infrared radiance observations, in combination with the MWR, have the potential to retrieve LWP with a higher accuracy than a MWR-only retrieval.