Abstract:This study examines the interannual variability of rainfall in western equatorial Africa and its links to sea-surface temperatures (SSTs). Five geographical regions within the latitudes 10°N-5°S are delineated for the analysis. The links to SSTs in the tropical Atlantic, Pacific and Indian Oceans are examined via seasonal composites of wet and dry years and via linear correlations.The results show that interannual variability is extremely complex in this region and that several factors govern it. The most important include SST anomalies along the Benguela Coast, a general warming or cooling of the tropical oceans, Atlantic SSTs specifically, the contrast between the Atlantic and Indian Oceans, and the Pacific El Niño-Southern Oscillation (ENSO). These factors differ seasonally. In much of the region, rainfall variability is linked to the Pacific El Niño and the western Indian Ocean early in the year, but to the Atlantic during the boreal summer months. The Indian Ocean again becomes important in late summer/early fall.The role of the Atlantic appears to be the modulation of the north-south excursion of the Intertropical convergence Zone (ITCZ). Hence the polarity of the SST/rainfall association depends on location. The association between SSTs near the coast and rainfall is positive if the influence is direct but can be positive or negative for indirect influences. An opposition between the Indian and Atlantic Oceans appears to displace convection in an east-west direction.Our results suggest several generic conclusions concerning the link between SSTs and continental rainfall. One is that the influence of the three oceans is seasonally dependent. The impact of a specific SST anomaly is also seasonally dependent. The same SST pattern may enhance rainfall in one season, but reduce it in the following season. Finally, the SST/rainfall associations are generally not symmetric. That is, the factors producing wet conditions are not the reverse of those producing dry conditions. In order to understand these associations, the underlying mechanisms via the general atmospheric circulation must be determined.

The diurnal variation of snow precipitation in the west coastal area near Wakasa Bay has been investigated for two winter seasons of 2001 and 2003 using surface precipitation radar, rain gauge, and satellite infrared data. Radar reflectivity-derived precipitation intensity shows a clear diurnal cycle within Wakasa Bay for both years, although the cycle is clearer in 2001 than in 2003. The precipitation maximum occurs in the early morning, and the minimum occurs in the evening. Using radar data collected during January 2003, the precipitation diurnal cycle within the Bay is compared with three nearby regions: offshore-open water region to the north, inland-land region to the northeast, and inshore-coastal region to the northeast. It appears that all the other three regions have the precipitation maximum during the day and the minimum during night, although the diurnal variation inland (over land) is very small. Additionally, in the offshore region, there exist two precipitation maxima and minima during a 24-hour day. Analysis of precipitation data from AMeDAS (Automatic Meteorological Data Acquisition System) rain gauge stations basically agrees with the radar observations. It shows that a similar diurnal cycle, found in the radar data within Wakasa Bay, can also be found for coastal stations, although the different patterns are shown depending on the location and time period in 2003. When averaging data from both January and February during 2001 and 2003, the diurnal cycle tends to be smoothed out. Cloud top temperature and cloud fraction derived from satellite infrared data do not show a clear diurnal cycle within Wakasa Bay, nearby inshore and inland regions, although the decreases of brightness temperatures are seen around noon in the offshore region for both cases of 2001 and 2003. Analysis using collocated satellite and radar data indicates that cloud top temperature has little skill in reflecting surface precipitation for the winter convective clouds associated with cold air outbreaks. Finally, possible causes of the diurnal variation are discussed, including local land-sea breeze and mountain wind effects, as well as the radiative cooling effect.