Identifying early morning fog and low cloud predictors for northern parts of the Northern Territory of Australia, utilising satellite-derived data

Abstract

Strong fog and low cloud predictors are identified for northern parts of the Northern Territory during dry-season months. A large-scale morning fog and low-cloud climatology was derived from geostationary-satellite imagery. This was statistically compared with meteorological station data from the previous afternoon. This approach utilised point location data to forecast the probability of occurrence of fog and low cloud over a larger area. Therefore, forecast domains are regional and not site specific. The satellite data indicated that during the early months of the dry season the bulk of the fog and low cloud occurred over eastern parts, whereas during the later months it was more widespread over the total area. Mean sea level pressure at central Australian locations and gradients of mean sea level pressure across northern Australia are good fog and low cloud predictors for southern and western parts of the region. Low-level moisture at locations along and inland from the Gulf of Carpentaria are better predictors for eastern land areas. This is explained in terms of the direction and strength of the easterly trade winds and the modification when these pass over the Gulf of Carpentaria. The total data was filtered in order to identify Australia-wide mean sea level pressure patterns corresponding to particular fog and low cloud distributions over the target area. Finally, predictor testing on an independent data-set showed best results for the eastern land areas where fog and low cloud was observed most frequently.