As Honorary Research Associate in the Department of Civil and Environmental Engineering,Faculty of Engineering, Li-Pen continuously collaborate with researchers from Imperial College London and supports suppervison of PhD and MSc projects.

Assistant professor in Computer Aided Engineering (CAE), with research focuses on spatial and temporal rainfall modelling and forecasting using advanced statistical theories.

Design and implement operational catastrophe (CAT) models for index-based micro-insurance products for natural hazards across Latin America area.

The consultant projects that I have been or am working on are mainly related to the setup of flood early warning system and statistical rainfall analysis. Some projects include:1. Development of a data-driven based rainfall and flood forecasting system for Birmingham city: This project is funded by EU (FloodCitiSense). My tasks include set-up and testing of a network of low-cost, IoT-enabled rain gauges over Birmingham city, and development of a data-driven urban (pluvial) flood forecasting tool for Birmingham city.2. Development of an inflow and flood forecast system for the operation of the Nam Ngiep 1 Hydropower Plant (HPP) in Laos:This is a consultant project with Artelia (France) and it is funded by NAM NGIEP 1 Power Company in Laos. My task is to co-develop a meso-scale rainfall and flood forecasting system with experienced meteorologist in Malaysia and FondUS technology Taiwan. The system will combine GFS forecasts with satellite precipitation products (IMERG and GSMap) and local rain gauge records in real time to provide reliable forecasts. 3. Development of real-time peril calculation tools (XSR, Drought and EQ) for micro-insurance company in Central America:This project is funded by Micro insurance Catastrophe Risk Organisation (MiCRO). My task was to develop calculation engines that allow real-time XSR, Drought and EQ calculation over Central America area.4. Development of integrated flood forecasting and river monitoring (iFFRM) with DMSS and warning system for Muar river basin in Malaysia:This project was funded by the Government of Malaysia. My task was to develop a gauge-based radar data processing routine that is now running operationally to provide radar QPEs in real time for the Sungai Muar catchment area.

This position is funded by the EU FloodCitiSense project. I am responsible for the technical implementation of a data-driven urban flood forecasting system for a pilot catchment in Birmingham (UK) based upon the Urban Living Lab methodology. This has entailed: - Organisation of co-creation workshops with local stakeholders and project partners.- Co-development of an IoT-enabled rainfall sensor data platform based upon open source istSOS software, Arduino and MQTT technologies.- Development and testing of various data-driven modelling approaches for urban flood forecasting.

This position is funded by the Plurisk project (http://www.kuleuven.be/hydr/plurisk). My task is twofold. The first one is to further develop radar-raingauge data merging techniques especially for urban-scale hydrological applications. The second is to improve fine-scale estimation of storm (wind) motion using advanced image processing (cell-tracking and optical flow) techniques. The resulting methods for motion estimation are being applied to urban rainfall nowcasting and design rainfall generation.

The consultant projects that I was involved in are:1. Estimation of fine-scale rainfall extremes for the design of nuclear power plants:This project was funded by EDF Energy. Based upon the extreme value theory and the Fractals, a methodology has been developed, providing robust estimate of 15-min rainfall extremes (up to 10,000 years return period).2. Reconstruction of a 2009 summer storm in Maida Vale, London:This project was funded by MWH, aiming for the reconstruction of the historical storm event that flooded the Maida Vale area (in Central London) in June 2009. A methodology to merge radar and rain-gauge rainfall data was developed to better reproduce the local rainfall peaks of storms, which is critical for urban-scale hydrological applications.3. Development of a spatial and temporal rainfall generator:This project was funded by MWH and Thames Water (UK). A C++ software tool was developed to sample storms events with spatial and temporal variations, of which the statistical features were calibrated using long-term radar rainfall images.4. Implementation and maintenance of the backup system of FEWS (Flood Early Warning System, developed by Delft Hydraulics) Live System in Taiwan:This project was funded by Water Resources Agency, Ministry of Economic Affairs, Taiwan. I set up the backup system for the operational FEWS Live System using a number of Open Source software (e.g. MySQL, Postgre SQL and Apache Tomcat), and produced the guideline for the installation.

This position was funded by the EU Interreg IVB RainGian project (http://www.raingain.eu/). My task was to develop and to implement advanced radar signal processing techniques for the UK Met Office (UKMO), aiming for improving the quality and the resolution of currently operational C-band radar rainfall data. A number of C/C++ signal processing routines were developed (based upon oversampling techniques) during this period, and have been installed on the Cyclops (i.e. the UKMO radar signal processing system). In addition, in order to shorten the computational time, one of the routines was further implemented under the OpenCL framework to be able to execute in parallel on GPU.

During my PhD programme, I also worked as a research postgraduate in the Flood Risk Management Research Consortium: Phase 2 (FRMRC2) project (funded by EPSRC). My task was to develop a stochastic rainfall downscaling model that enables the generation of urban-scale rainfall estimates. A number of C++ routines were developed and implemented to conduct spatial and/or temporal rainfall downscaling, based upon the theory of multifractals. In particular, the spatial downscaling routine has been further implemented as an OGC Web Processing Service and is now being tested within ICL web domain.

This position was funded by the Ministry of Education of Taiwan under the Programme for Excellence Research Team. My task was to developed a simulation tool for numerically modelling behaviours of fluid-solid interactions by integrating existing theories and verified the associated results by comparing them with experimental results. The prototype of this tool was developed during this period by coupling the open source lattice Boltzmann (OpenLB) library with an in-house discrete element modelling (DEM) system.