Assessment Of The Influence Of Astronomical Parameters On The Skill Of Rainfall Forecasting In East Africa.

Abstract

Seasonal forecasts generated in East Africa have mainly used the SST-based models in the last two decades with challenges of poor forecasting skills. In particular, the method of forecasting rainfall extremes has been too general and that these extremes occur much more frequently than forecasted. And furthermore, no forecasting model in use in the region provides the temporal variation or the intra-seasonal-to-interannual variability of rainfall. In addition to the low skills, the traditional Indigenous Knowledge (IK) forecasting methods are falling out due to climate change. One component of the IK, the astronomical observations, is viewed with a lot of scepticism and is considered as a non-science, therefore, inhibiting its application in science-based forecasting and research. This study focuses on astronomical observations in its objective which is to determine the influence of the orbital parameters of planets and the moon on the weather and climate patterns in East Africa. Results generated show that Saturn, Jupiter, Venus and Mars have a relationship with rainfall but at different levels. Both MAM and OND in all zones seem to show a variation from year to year that indicates strong astronomical influence in most cases. We also note that rainfall characteristics during two similar celestial phases but which occur at different times of the year are different, however, rainfall characteristics associated with the same observed phase and in the same month or period were found to be nearly the same. To get to the same phase in the same month of the year, would take Saturn 30 years, Jupiter 12 years, Mars 15 years and Venus 8 years giving rise to what is refered to here as Saturn Rainfall Cycle, Jupiter Rainfall Cycle, Mars Rainfall Cycle and Venus Rainfall Cycle respectively. That means that the East African rainfall varies in cycles of 8, 12, 15 and 30 years. The rainfall cycles are easily determined by use of their key phases and can be predicted by use of astronomical calculations with little error and with sufficient accuracy way ahead of time. Further, by using historical information, it was found out that severe climate extremes occur during the conjunctions of both Uranus and Neptune where Uranus takes ~83 years and Neptune takes ~163 years to orbit to the next conjunction. These periods, now called Uranus rainfall cycle and Neptune rainfall cycle respectively, coincide with the variation of severe extreme events in the study area. We can attribute those variations to the two planets’ orbital motions. vi The probabilistic models developed here use probability of occurrence or exceedance and have five categories; “Extreme Low”, “Below Normal”, “Normal”, “Above Normal”, “Extreme High” and “Phenomenal”. Using the probabilities of occurence on 2018 rainfall seasons, a qualitative verification process indicated relatively high probability values of up to 67% under “Above Normal” and “Extreme High” for MAM 2018 forecast in areas that mainly fall in the highlands East of the Riftvalley, while the period OND indicated high probability values of up to 83% under the category “Below Normal”. The season MAM 2018 was extremely wet and OND 2018 was extremely dry which means the probabilities had captured the extremes as projected. Generally, from the results, it was found out that the planets have a relationship with the East African rainfall. Each one of them showed a certain level of contribution to the variation of monthly rainfall with the Planet Saturn indicating the biggest influence. The moon had relatively little influence to the monthly rainfall variation compared to the planets. The phases of the planet can be hindcasted back in time to allow a dependable determination of rainfall variation of the past. In general, the use of these astronomical phases can be used to generate past and future climate scenarios in the region that can add useful body of knowledge to climate science that can be integrated in scientific reports like the IPCC Assessment Reports.