Overview

The world population is increasing and many cities (particularly in developing countries) are expanding rapidly – often with little or no planning and associated infrastructure. At the same time, climate change predictions suggest that many parts of the world will experience increased frequency of extreme rainfalls. This make cities more vulnerable to pluvial flooding when rainfall and runoff rates exceed urban drainage capacities. This can lead to damage to buildings, traffic disruption and fatalities. Models are now routinely used for pluvial flood prediction and risk assessment, with ever-increasing spatial and temporal resolution. 

However, these models need accurate and timely measurements of surface precipitation: radars (e.g. the Met-Office radar network) and telemetry-linked rain gauges have been traditionally adopted for rainfall estimations. However, radar is known to suffer from several issues including occultation by clutter and degradation of resolution and decrease in accuracy with range.  Furthermore it is not available in the majority of developing countries. In many parts of the world the density of surface precipitation gauging networks is rapidly declining and only few gauges are telemetrically linked. One alternative is to use the received signal level data from the enormous number of microwave links used in commercial cellular communication networks and/or from satellite TV antennas. Commercial link networks cover large parts of the land surface of the earth and have a high density, particularly in urban areas.

Satellite TV dishes have even greater potential in terms of crowd-sourcing precipitation data. Rain induces attenuation in the radio signals that propagate from a transmitting to a receiving antenna (for cell-phone links) or from the satellite to the TV dish. The signal attenuation between transmitter and receiver is almost linearly related to the path-averaged rainfall intensity. This attenuation can be calculated from the difference between the received powers with and without rain. This concept is now well-established for cell-phone links [e.g. 1,2] and is currently under investigation for satellite TV downlinks. However, its exploitation as an input in urban hydrological models [3] and the development of flash-flood early warning systems in (mega)cities has yet to be attempted.

Flooded roads in Durban, South Africa, October 2017.

Methodology

The PhD student will get acquainted with state-of-the-art methodologies for mapping precipitation by exploiting the attenuation in cell-phone signal. These techniques will be applied over a pilot study area (e.g. Nairobi, Kenya). Precipitation estimates will be validated using other precipitation data (gauge and satellite-based) and a general space-time precipitation data product will be developed.

The spatial and temporal dynamics of precipitation will be used to drive an established 2D urban hydrological model capable of predicting pluvial flooding in near real-time (i.e. with results in seconds or minutes). The methodology will be applied to a city with known risks to pluvial flooding and with available microwave link data. The model outputs will be compared with historical and current observed data on the extent and depth of inundation in order to evaluate the suitability of the derived rainfall products, compared with models driven by precipitation datasets obtained using traditional methods.

Training and Skills

This project offers an excellent opportunity to develop and apply novel techniques in rainfall remote sensing, to work with the latest generation of 2D urban hydraulic models and to acquire skills in the processing of complex spatially referenced datasets. The student will be trained in a wide range of topics including hydrological and hydraulic modelling, spatial data processing and precipitation remote sensing. Applicants should have a science or engineering degree with a strong mathematical component.  Knowledge of hydrology and or hydraulics would be beneficial. Programming skills in matlab/idl/Python/C/Java/C++ and knowledge of signal propagation and numerical modelling would also be beneficial.

 

Timeline

Year 1: The student will explore the potential of microwave cell-phone links for rain detection and quantification. The student will develop methodologies to optimally merge rainfall products available from different remote sensing techniques and from in situ measurements (rain gauge data).

Year 2: The student will use the precipitation fields derived from the cell-phone signal changes as inputs to urban pluvial flooding models (e.g. ISIS FAST [3]). This will require acquisition and quality control of detailed topographic and drainage infrastructure data for the areas of application.

Year 3:   An early warning system will be developed using the combination of urban hydraulic modelling and precipitation data products derived from cell-phone signals. 

 

Partners and collaboration (including CASE)

A close collaboration between the Leicester Earth Observation Science group and the School of Geography, Geology and the Environment  is at the core of this project. Dr Battaglia is a cloud and precipitation microwave remote sensing expert with more than 60 co-authored papers and more than 15 years of experience in the field. Dr Whelan is a hydrologist with extensive experience in the development and application of hydrological models. The student will benefit from technical support and use of facilities in the Leicester Space Research Centre.

Further Details

For further information please contact Dr. A. Battaglia, email:ab474@le.ac.uk