This project will use newly developed regime-conditioned sub-seasonal precipitation forecasts to design and evaluate an early warning system (EWS) of extreme events over Southeast Asia. The EWS skill will be evaluated for different thresholds, temporal, and spatial scales. Its skill will also be compared with an EWS derived from raw precipitation forecasts.
Department: Meteorology
Supervised by: Oscar Martinez-Alvarado
Southeast Asia is frequently impacted by severe weather that causes damage and loss of life and property, and stalls the regional development. Within the framework of the Newton Fund Weather and Climate Science for Service Partnership (WCSSP) programme, a group of scientists at Reading have developed a set of circulation regime-conditioned forecasts of extreme precipitation for Southeast Asia that show encouraging results, with positive levels of skill up to around 20 days ahead. Reliable severe weather forecasts are fundamental for the design, preparation, and deployment of relief responses. However skilful, the standard format of model-derived probabilistic forecasts is not easily accessible to users. To transform these forecasts into a useful tool for decision makers, this project aims to design an early warning system (EWS) for extreme precipitation events that could be used to ‘trigger’ hypothetical emergency responses at different lead times. The student involved in this project will make use of the existing forecast set and conduct sensitivity studies to design an EWS for different trigger probabilities, event definitions and spatial and temporal scales. The skill of such EWS will also be compared with a similar system derived from the raw precipitation forecasts. We aim to have input from local WCSSP partners (Philippines, Indonesia, Malaysia and Vietnam) in the design of the EWS, in terms of risk and exposure parameters (event definition, relevant regions), but also in identifying key lead times for emergency response. We also aim to get their feedback in terms of the graphical representation of the products.
The first week of the project the student will familiarize with the forecast set and some sample code to open and process them. They will also review relevant literature. Over the following 2-3 weeks, the student will conduct tests implementing an EWS for different forecast probabilities and precipitation thresholds; and explore the use of different skill metrics to evaluate the adequacy of the hypothetical response (for example, in terms of hit and false alarm rates). During weeks 4-5 we will implement it as a preliminary product and share it with local partners for feedback. We will aim to test it for previously identified case studies in addition to the overall skill assessment. The student will finally develop a written description of the methodologies and the results. This stage will take around 1 week.
- Experience with a programming language such as python, MATLAB, or R - Some knowledge of basic statistics will be beneficial - Interest in the graphical representation of complex science results in a format accessible to policy makers and the public
- The project will provide enhanced experience in programming and managing large datasets - The student will gain experience working on a Linux environment - If the results allow it, the student might have the opportunity to contribute towards a publication or conference presentation - The student will gain experience in climate services research and impact-based forecasting - They might have the opportunity to participate and present in meetings with the local partners (meteorological services in the region).
Department of Meteorology, computer lab (subject to CV-19 restrictions and University guidelines)
Start and end dates to be agreed. Flexible working hours, but as a reference: 9am to 5pm, 1 hour lunch break
Monday 13 June 2022 - Tuesday 20 September 2022
The post will be advertised centrally on the UROP website between 21st February and 4th April 2022. Students should submit their CV and Cover Letter directly to the Project Supervisors (Oscar Martinez-Alvarado: o.martinezalvarado@reading.ac.uk / Paula Gonzalez:p.gonzalez@reading.ac.uk). Successful candidates will then be invited to interview.