Augustine Kiptum

I am climate scientist with expertise in weather and climate forecasting spanning more than 10-year having worked at the Kenya Meteorological Department (KMD) forecasting section and currently as a PhD Researcher at the University of Sussex, UK.

My main task at KMD were application of numerical and dynamical weather and climate models to issue forecasts. Locally, at the forecasting section, with other climate forecasters, we were able to initialize and run WRF model at medium range (7-day lead) that could feed into the seven day forecast as well as input for flood forecasting models at the Flood Forecasting and Warning Centre.

In a number of occasions, together with other climate scientists in the IGAD region participated and make seasonal forecasts during the Climate Outlook Forums(COFs).

Currently, I am in my final year as PhD student at the University of Sussex, UK my main area of research has been the understanding potential application of long lead numerical weather prediction products from ECMWF's Integrated Forecasting System(IFS) in the initialization of spatially distributed hydrological model-LISFLOOD forecasting in Kenya focusing on two flood prone basins of Tana and Nzoia. Ideally, the main work here is to assess skill of GloFAS at sub seasonal time scales by comparing with observed streamflow and available flood impact datasets from Kenya Red Cross and international Sources(Dart-Mouth Flood Observatory and EMDAT, DisInventer etc). Through this work, I have submitted one paper and accepted at the Journal of Flood Risk Management and available here: DOI: 10.1111/jfr3.12884 . I have one paper which is currently in the final stages of submission for review.

Through my PhD research, I have also developed strong technical and analytical skills on data analysis through expertise in Python, R Climate Data Operators(CDO). In addition, I have developed expertise in using the Google Earth Engine where my work has been looking into the flood risk mapping using Sentinel products and then applying Machine Learning (ML) techniques(Random Forest[RF] Classification and others) in the classification of flood prone areas.

I have submitted a review paper in the Journal of Flood Risk Management which has been accepted.

Kenya Meteorological Department(KMD) and University of Sussex