Near real-time correction of flood forecasting using high-resolution satellite data for multi-hazard risk assessment in lowland tropical regions.

Lead Research Organisation: Cranfield University
Department Name: School of Water, Energy and Environment

Abstract

Flood modelling and forecasting are essential tools to inform infrastructure and emergency planning. Accurate forecasts, though, are difficult to achieve, even in developed countries with decades of experience and detailed topographical and hydrological datasets for calibration and validation, as demonstrated by the recent Cumbria floods. Forecasting is even more challenging in large tropical regions, which have limited data availability (e.g. few river gauging stations), and modest meteorological forecasting capabilities. More worryingly, their large human populations, economically-important industries (e.g. oil/gas, agriculture), and ecologically-important habitats mean that flooding is connected to multiple other significant risks.
This PhD project will attempt to overcome some of these challenges by (1) using new near real-time, high resolution satellite datasets to improve the medium- and short-term flood risk assessment generated by probabilistic ensemble flood forecasting for data-poor tropical regions, and (2) applying the flood model to the assessment of flood-induced pollution risk. The case study for the project will be the Mexican State of Tabasco, which occupies a large, low-lying, topographically-complex area that experiences flooding from several large rivers (e.g. Grijalva-Usumacinta systems, 1,911 km long network, 128,390 km2 catchment area) which are affected

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
NE/M009009/1 05/10/2015 31/12/2022
1819435 Studentship NE/M009009/1 03/10/2016 02/10/2019 Charles Mazivanhanga
 
Description Given the rapidly increasing availability of satellite datasets and the power of computers, the study has shown that it is feasible to use high resolution distributed hydrological modelling and satellite data to accurately represent the highly variable catchment conditions and processes and generate reliable flood simulations in data-scarce large tropical catchments

The study expanded the application of statistical scaling relationships between peak discharge and drainage areas in nested tropical catchments and used the relationships to estimate and correct the timing and magnitude of peak floods.

In order to improve the reliability of model results, the study managed to disentangle the rainfall and parameter uncertainty and quantify individual uncertainties taking into account the complex interactions and effects on model outputs.
Exploitation Route Given the rapidly increasing availability of satellite datasets, computer power and distributed models, it is possible to meet the challenge of providing timely and reliable flood simulations in large data-scarce tropical regions where millions of people are exposed to risks of flooding.
Sectors Digital/Communication/Information Technologies (including Software),Environment,Other
 
Description Stakeholder presentations for CONAGUA, Department of Emergencies and City Council resulted in great interest in the integrated / catchment-wide flood modelling approach used in this study. The Villahermosa City council went as far as to requesting consultancy fees for flood maps with possible flood scenarios for decision-making for its city land-use planning.
First Year Of Impact 2019
Sector Communities and Social Services/Policy
Impact Types Policy & public services
 
Title Shetran model auto calibration tool 
Description Instead of using manual calibration, developed an automatic calibration tool for the SHETRAN hydrological model using OSTRICH (Optimization Software Toolkit for Research Involving Computational Heuristics), a model-independent optimisation toolkit 
Type Of Material Improvements to research infrastructure 
Year Produced 2018 
Provided To Others? No  
Impact The calibration of nested catchment models was conducted using the Dynamically Dimensioned Search (DDS) algorithm as implemented in OSTRICH, which is a model-independent calibration and optimisation tool. A faster and more efficient calibration approach than the manual calibration of the model. 
 
Title A large scale distributed hydrological model 
Description A large scale distributed hydrological model with a 2 x 2 km grid cell resolution was developed using the SHETRAN hydrological model and run enforced with a 3-hour satellite input rainfall data able to simulate river discharge in a data-poor tropical Grijalva catchment area (57,958 km2) 
Type Of Material Computer model/algorithm 
Year Produced 2019 
Provided To Others? No  
Impact Given the rapidly increasing availability of satellite datasets and the power of computers, this study developed a fully distributed hydrological model integrated with high-resolution satellite datasets to improve the simulation of river flows in a large data-poor lowland tropical catchments. The results show that accurate representation of large scale catchment conditions and processes using high resolution fully distributed hydrological models integrated with high spatial (< 2 km) and temporal (3 hr) resolution satellite datasets can provide reasonable levels of flow simulation accuracy. In this way, the present study contributes to solving the current problem of flood forecasting in data-poor tropical catchments where millions of people are exposed to flooding. 
 
Title Multi-scale nested catchment hydrological models 
Description Six multi-scale nested distributed hydrological models all with 1 x 1 km grid cell resolution were developed using the SHETRAN hydrological model. All were run enforced with 1, 2 and 3-hour satellite input rainfall data and were capable to simulate river discharge in a data-poor tropical La Sierra nested catchment sizes ranging from 128 to 6 743 km. 
Type Of Material Computer model/algorithm 
Year Produced 2018 
Provided To Others? No  
Impact The six multi-scale nested distributed hydrological models were used in the application of statistical scaling equations for estimating the magnitude of peak discharge in La Sierra catchment 
 
Title Statistical scaling equations 
Description Five statistical scaling equations were developed for estimating flood parameters and peak discharge in la Sierra catchment and sub-catchments. 
Type Of Material Computer model/algorithm 
Year Produced 2018 
Provided To Others? No  
Impact Five statistical scaling equations developed provided a framework for estimating flood parameters and magnitude peak discharge for monitoring the downstream generation of flooding in la Sierra catchment and are complementary tools for verifying and refining the magnitude of the peak discharge simulations generated by the distributed hydrological models. 
 
Description Newcastle University 
Organisation Newcastle University
Country United Kingdom 
Sector Academic/University 
PI Contribution I have been involved in consulting experts/lecturers at Newcastle University on modelling work and for support, if I find any problems with the construction of a satellite-based distributed hydrological model.
Collaborator Contribution Providing expert advice and providing specialist training courses on flood modelling and forecasting.
Impact Building a large scale satellite-based hydrological model for Grijalva basin Mexico. I had three modelling experts that I work with.
Start Year 2016
 
Description Universidad Juárez Autónoma de Tabasco, Mexico 
Organisation Juárez Autonomous University of Tabasco
Department Faculty of Biological Science
Country Mexico 
Sector Academic/University 
PI Contribution Throughout my fieldwork activities in Mexico, I have worked directly with my Industrial Supervisor who facilitated my data collection, stakeholder meetings, seminars and workshop presentations. I participated in leading meetings, seminars and conferences, and presented research progress and findings.
Collaborator Contribution The studentship is co-funded through the Data, Risk and Environmental Analytical Methods Doctoral Training Centre at Cranfield University and Juárez Autonomous University of Tabasco (UJAT). The Funding from UJAT cover costs of travel and subsistence for research visits to Mexico. The funds from Mexico also cover annual research trips for the Academic Supervisor to visit UJAT and the industrial Supervisor visit to Cranfield University
Impact So far three research trips to Tabasco State Mexico by the student has taken place. This includes a trip by one Cranfield Supervisor to visit UJAT and one visit done by the industrial Supervisor to Cranfield University.
Start Year 2016
 
Description DREAM CDT Symposiums every year 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact The annual DREAM-CDT (Data, Risk and Environmental Analytical Methods-Doctoral Training Center) Symposium are industrial seminars that provide a great opportunity for doctoral researchers to come together to present their latest research findings and to hear from industry leaders such as Microsoft, Google & ESRI and others.
Doctoral researchers from Cranfield, Cambridge University, Newcastle University and Birmingham University come together to discuss their latest research results on how to use ' big data ' to reduce environmental risks and ask questions. Researchers also hear presentations from keynote speakers and a number of industry delegates and exchange ideas at each event.
Year(s) Of Engagement Activity 2017,2018,2019
 
Description Seminar presentation, La Universidad Juárez Autónoma de Tabasco (UJAT), Villahermosa city, Tabasco, Mexico 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Undergraduate students
Results and Impact I gave presentations and attend meetings with academic staff and students at the Universidad Juárez Autónoma de Tabasco (UJAT). This includes consultations with scientist and engineers from CONAGUA and the Department of Civil Protection and also networking with other key stakeholders in flood risk management in State.
Year(s) Of Engagement Activity 2016,2017,2018
URL http://www.ujat.mx/Noticias/Interior/24757
 
Description The National Water Commission CONAGUA presentation 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Presentation to update the National Water Commission CONAGUA of Mexico on research progress, findings and data requirements.
Year(s) Of Engagement Activity 2018,2019
 
Description Villahermosa Local Council Seminar presentation, 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Third sector organisations
Results and Impact Consultative stakeholder meetings on flood risk problems, needs and requirements in Villahermosa City, Tabasco, Mexico. Update on research progress and outcomes. Huge interest was shown in the research outputs. The Council asked for consultancy fees for the production of its own flood risk maps that can allow land-use planning in Villahermosa city.  
Year(s) Of Engagement Activity 2018,2019