The first major thrust of this project was to improve our knowledge and understanding of the nature and variability of African easterly waves (AEWs). There were several major outcomes from this part of the work that highlighted the presence of significant variability in AEW structures and intensity. Based on analysis of European Center for Medium-Range Weather Forecasting (ECMWF) reanalysis data combined with satellite-derived brightness temperature we were able to highlight two contrasting types of AEW behavior. One had confined convection and circulation features, mostly south of 15^oN, that shares many characteristics with AEWs analyzed and discussed in the literature. In contrast, a second type of behavior was also identified that had less familiar characteristics. The most striking feature of this AEW was its broad meridional extent, which resulted in interactions with both the subtropics to the north and the equatorial region to the south. Indeed, the interactions in the equatorial region resulted in a previously unknown hybrid structure that shares characteristics with a large-scale Mixed-Rossby-Gravity wave (MRG) - a well-known wave that propagates zonally around the equator. The origins of the MRG wave over tropical Africa were explored and shown to be associated with extratropical weather systems in the Southern Hemisphere moving into the tropical Atlantic. This result has consequences for the predictability of the AEW-MRG interactions on timescales of days to weeks. A broad survey of AEW intensity characteristics was carried out using an automatic tracking technique. This highlighted the presence of considerable variability in AEW intensity and its causes. Most notably, AEWs with stronger circulations over West Africa generally arise in association with larger accumulated rainfall totals during the previous days. This work provided training for a PhD student in the dynamics of tropical waves including the statistical tools needed to carry out the analysis. PhD student Yuan-Ming Chen earned his PhD based on this work.

The second major thrust of this project was to understand the factors that modulate our ability to predict the evolution of AEWs.  This was accomplished by analyzing AEW forecasts from the ECMWF ensemble prediction system.  This system provides multiple forecasts that are initialized at the same time; therefore, if the simulations have different forecasts of the AEW, it is indicative of a lack of ability to predict that feature.  This analysis was carried out over two periods (2007-2009 and 2011-2013), which were characterized by significant numbers of AEWs, but different versions of the ECMWF model.  AEW position and intensity forecasts from 2007-2009 were less skillful than forecasts from 2011-2013. One of the reasons for the less skillful forecasts from 2007-2009 is related to how the two versions of the model simulate convection.  In the 2007-2009 version, the model produces excessive rainfall and convection relative to satellite-based observations, which in turn causes the AEW to become too intense in the forecast.  By contrast, the 2011-2013 forecasts show a weaker relationship between convection errors and AEW errors.  The second phase of this work investigated whether certain locations or situations are associated with less predictable AEWs.  Less predictable AEWs are characterized by higher relative humidity in between 2-5 km above the ground, which in turn results in more variability in rainfall and convection, and hence in the AEW itself.  By contrast, the variability in AEW forecasts does not appear to be sensitive to non-convective processes that could also lead to the intensification of AEW. This work provided training in atmospheric prediction and dynamics for Travis Elless, who earned a Ph.D. through this project.

Last Modified: 08/30/2019
Modified by: Christopher Thorncroft