A cloud‐detection scheme for use with satellite sounding radiances in the context of data assimilation for numerical weather prediction

Abstract

A scheme for detecting cloud-affected radiances is described. The method is used to determine the probability of cloud-free conditions given the observations and the prior knowledge we have about the atmosphere from a numerical weather prediction (NWP) model. This is achieved using a likelihood method. It combines the strengths of some alternative methods (e.g. comparison of infra-red and microwave channels sounding the lower troposphere and comparison of infra-red window channels with sea surface temperature) in a powerful and flexible method. It is powerful because it uses different types of information simultaneously. It is flexible because it makes no assumption about which instrument is being processed, or what type of prior information ( M , climatology etc.) is used. Therefore, it can readily be extended to new situations and data types (e.g. Advanced TIROS Operational Vertical Sounder (ATOVS)). It is suitable for use on general cloud-detection problems, using combined microwave and infra-red data. It has been tested using TIROS Operational Vertical Sounder (TOVS) radiances. The new method has been compared with an alternative cloud-detection method tailored specifically for TOVS and has been developed to a level of robustness adequate for operational use. The new method gave very similar results to the alternative method, especially over the ocean. The differences that did occur have been investigated by comparing with cloud information derived from the Advanced Very High Resolution Radiometer (AVHRR). Both the alternative method and the new scheme were found to have deficiencies when dealing with very low cloud. Some cloud missed by the existing scheme is identified by the new scheme. Over land, cloud detection is more difficult. The two schemes disagree more often, but validation using AVHRR is also more difficult because of increased surface heterogeneity and more variable emissivity and surface temperature errors. The new method is therefore shown to perform at least as well as an alternative method in operational use, whilst gaining the flexibility required for future systems. The implications for ATOVS are discussed.

Bayesian cloud detection for MERIS, AATSR, and their combination

“…7.1, but this choice clearly depends on the target application and is independent of the Bayesian approach. Other applications, such as the construction of a cost function as described by English et al (1999), are also viable alternatives.…”

“…English et al (1999) used Bayesian probability with strongly dependent features to derive a cost function for a 1D-Var retrieval of cloudiness. The classification is based on the exploitation of microwave and infrared channels and, in addition, external data from NWP simulations.…”

“…Common approaches to cloud masking are hierarchies of thresholds (e.g., Rossow and Garder, 1993;Schlundt et al, 2011), complex statistical models (e.g., Murtagh et al, 2003;Gómez-Chova et al, 2008), or other Bayesian approaches (e.g., English et al, 1999;Uddstrom et al, 1999;Merchant et al, 2005;Mackie et al, 2010a;Heidinger et al, 2012). A classification scheme for Bayesian cloud masks which helps to clearly distinguish the various approaches to Bayesian cloud masking is introduced in Sect.…”

Assimilation of IASI at the Met Office and assessment of its impact through observing system experiments

“…If (Predicted BT -IASI BT) > 5 K the pixel is cloudy. The next test follows the Bayesian methodology of English et al (1999) which takes into account information from the IASI observation, collocated AMSU and the model profile. No IASI data are assimilated where one of these three cloud tests is failed.…”

Ready-to-Use Methods for the Detection of Clouds, Cirrus, Snow, Shadow, Water and Clear Sky Pixels in Sentinel-2 MSI Images

“…Such schemes are in general distinct for a particular instrument, although basic physical principles for similar spectral bands hold among instruments. Some examples from the relevant literature can be found in a variety of references [1,[17][18][19][20][21][22][23][24][25][26][27][28]. Cloud detection is the main focus for most of these references, while the aim of this study is the separation of all introduced classes.…”