Emeritus Professor Howard Wheater

Professor Wheater moved to Canada in October 2010, as Canada Excellence Research Chair (CERC) in Water Security at the University of Saskatchewan (UofS). In 2011 he established the Global Institute for Water Security, with some 80 Faculty and senior government researchers, and has created interdisciplinary teams to address the Institute’s key themes (Climate Change and Water Security, Land-Water Management and Environmental Change, Sustainable Development of Natural Resources, and Socio-Hydrology). An ambitious programme of research focusses on the 340,000 km² Saskatchewan River Basin (SaskRB) as a large-scale observatory. The CERC currently supports 6 Faculty, 7.5 Undergraduate Students, 29 Graduate Students, 16 Postdoctoral Fellows and 20 Research and Administrative Staff.

With CERC and CFI support, the SaskRB study includes hydro-ecological observatories in the key biomes of the Rocky Mountains, Boreal Forest and Prairies, including collaboration with the Canadian Space Agency and NASA, to support improved process understanding and modelling across multiple scales. Water resources simulation work has developed new methods of vulnerability analysis for complex water resource systems, and systems dynamic simulation tools for scenario-based economic analysis of water futures, supported by new downscaling tools for climate model outputs. Water quality research has focussed on hydro-ecological response to nutrient loading in the South Saskatchewan River (in particular Lake Diefenbaker), and the development of improved modelling tools for the assessment of the impact of agricultural Beneficial Management Practices. Dr Wheater has also chaired a Council of Canadian Academies Expert Panel that has addressed more generally the relationships between water and agriculture for Canada. Socio-hydrology research has included stakeholder engagement with experimental programmes, and a series of workshops to explore water security issues and concerns across the basin. The SaskRB study has been recognised by the World Climate Research Programme GEWEX (Global Water and Energy Exchanges) Project as a Regional Hydroclimate Project (in December  2012). An important extension of the scope of this research has been the 2013 NSERC award of $5million funding of a Changing Cold Regions Network for the assessment and diagnosis of environmental change in the interior of Western Canada, including the Mackenzie and Peace-Athabasca river basins, as well as the SaskRB. However, while publications are beginning to flow from this work (see selected list below), his major outputs over the last 6 years are primarily based on previous work, as outlined below.

Professor Wheater was until 2011 co-PI of the £14.1M UK Flood Risk Management Research Consortium research programme, funded by the UK Research Councils (EPSRC and NERC), national and regional governments, the Environment Agency of England and Wales, and the UK water industry. There is concern in the UK that changing rural land use and land management has increased flood risk, and major policy interest has arisen in aligning agricultural management practices with flood mitigation measures. Professor Wheater lead national research into the effects of rural land use and land management change on flood risk, and the potential for flood risk remediation. He secured additional funding for the second phase of this programme from the NERC FREE initiative.

A unique multi-scale experimental programme was developed to quantify hydrological effects of land use interventions at plot, hillslope, and catchment scale, working with a consortium of farmers in the headwaters of the river Severn in upland Wales. The data have supported a wholly new modelling methodology to quantify local scale effects of land management and catchment scale implications for flood risk. Highly detailed physics-based models of local scale interventions have been developed, and emulation modelling has been used to upscale the results. The research has been generalised to address a wider set of land use treatments and catchments, including the development of a new physics-based model of upland peatland management, conditioned on both local and surrogate data. Finally, new regionalisation approaches have been developed for ungauged catchment application using Bayesian conditioning on hydrological indices derived from national and international databases.

This research provided the first data and modelling methods to quantify these land use effects on flood risk for the UK, and is providing input to policy for national, devolved, and regional governments. The research has received much public attention, with visits from His Royal Highness the Prince of Wales and Welsh Assembly Government Ministers, and national radio interviews. It contributed to a government foresight study of land use futures, to Sir Michael Pitt’s Cabinet Office review of the 2007 floods in England and Wales, and to previous UK government Flood Foresight studies. Professor Wheater also provided oral evidence to the House of Commons 2007-2008 inquiry into flooding.

Professor Wheater secured funding for a £10M national research programme, supported by NERC, into the hydro-ecological functioning of lowland, groundwater-dominated catchments. The research was motivated by the need to develop the science base to support UK catchment management under the EU Water Framework Directive. It focussed on three catchments, two on the Chalk of southern England and one on the Triassic Sandstone of the Midlands. He won a national infrastructure personal grant of £2M for field experimental facilities to support the initiative, and in the subsequent programme he coordinated research on the Pang and Lambourn catchments, editing a special issue of the Journal of Hydrology to report the results.

His research focussed on the hydrogeochemical functioning of Chalk catchments, aimed at addressing nutrient pollution and nutrient management. He led a team comprising the British Geological Survey, NERC Centre for Ecology and Hydrology, Reading University, and Imperial College. His research included experimental and modelling studies of water flow and solute transport in the Chalk unsaturated zone (a fractured porous matrix). The research showed that nitrate pollution moved primarily through the fine porous matrix, with an associated ‘nutrient time-bomb’ of decades of historical pollution moving through the unsaturated zone and yet to reach the groundwater system. He developed a simplified catchment-scale management model to represent these processes and hence quantify effects of climate change and management options on future river nutrient concentrations. Professor Wheater subsequently advised the Environment Agency on the national assessment of Nitrate Vulnerable Zones, which now apply to 70% of England and restrict agricultural use of organic and inorganic fertilisers.

Groundwater flooding (i.e., surface emergence of groundwater) can be of long duration (weeks/months) and hence highly damaging to surface and subsurface infrastructure. Extensive groundwater flooding in England occurred in 2000-2001, which highlighted a national lack of knowledge, reporting, and tools for risk assessment and forecasting. Professor Wheater led a team funded under the NERC FREE programme to address this national need. Building on the LOCAR infrastructure above, the research has provided new insights into the groundwater system response under extreme conditions and has demonstrated a highly non-linear recharge response of the unsaturated zone, and the complexity of stream aquifer interactions. Limitations of current physics-based groundwater models have been demonstrated, and a programme of model development initiated. Simplified models were developed and tested, and have shown considerable power as potential risk assessment and forecasting tools.

Traditional flood design is based on assumptions of climate stationarity, with design hydrographs based on individual synthetic storm events. Several important limitations can be overcome by the use of continuous simulation hydrological modelling, but this requires appropriate precipitation inputs. A major research programme was funded by the UK government to provide stochastic rainfall modelling procedures to support a new generation of flood estimation methods.

At the hourly time scale, building on previous work, single site Poisson process models were evaluated for application to lumped catchment modelling, including detailed analysis of model identifiability, recommendations were made for national application, and the models provided for an integrated analysis of rainfall-runoff model performance. An extensive radar rainfall archive was created and used in the further development of continuous space-time Poisson process models, also at hourly time scale.

At the daily time scale, Generalised Linear Models (GLMs), previously developed by Chandler and Wheater to represent daily space-time rainfall fields, were refined and a web-based software package developed. In parallel, a PhD student developed and tested a new methodology for temporal downscaling of the GLM-based daily spatial temporal modelling using Poisson-cluster models, thus providing hourly precipitation fields for input to hydrological models.

In a second contract, a new methodology was developed to simulate space-time rainfall under climate change scenarios. Given the unreliability of precipitation products from Global and Regional Climate Models (GCM/RCM), GLMs were used to develop stochastic simulation models conditioned on the more reliable GCM/RCM outputs, such as temperature, pressure, and humidity. Applied to generate ensembles for a range of GCMs and RCMs, they have an important advantage over most alternative disaggregation schemes in that they represent spatial structure as well as temporal properties. With support from Imperial’s $25M Grantham Institute for Climate Change Research, the GLM methodology developed above was tested further, to evaluate in more detail extremes and persistence, and a new methodology developed to generate evaporation time-series, also using GLMs. This provides a complete weather generator for assessment of climate change hydrological impacts. Climate change impacts have been produced for a set of UK catchments, demonstrating the effects of the regional variability of the climate change signal, the important effects of catchment characteristics, and limitations of current methods in drought simulation. This work continues, with Canadian application to precipitation extremes and evaporation in the Prairies, and application of the BERMS fluxnet site data in the evaluation of climate model performance.

In previous work, a rainfall-runoff modelling toolkit had been developed that allowed the rapid testing of alternative model structures, and the use of advanced methods of stochastic analysis of model structure and performance. A major goal of the toolbox was to support research into the modelling of ungauged catchments – subsequently a key aim of the IAHS PUB (Prediction in Ungauged Basins) initiative. The book by Wagener at al., describes the toolboxes, sets out a theoretical framework for addressing the ungauged catchment problem, and provides a preliminary analysis of UK catchments. The book has been popular – a second edition is in preparation. In subsequent PhD research, following extensive model testing using a national database of 150 catchments, new regionalisation methods were developed for the UK, initially based on regression on catchment characteristics. A new approach to regionalisation was also developed, based on donor catchments providing multiple parameter sets, weighted in a Bayesian framework according to catchment similarity and donor catchment performance.

There is an important gap in understanding the spatial controls on catchment hydrological response, including the role of precipitation.  A detailed data set has been developed for the River Lee catchment in the UK, including radar rainfall data, and dense raingauge and flow monitoring, and extensive analysis has been performed to understand the controls on catchment runoff response, including precipitation properties, catchment characteristics, and scale. This has led to recommendations for raingauge densities for UK application.

Arid zone rainfall is often dominated by spatially-localised convective rainfall, which creates difficulties for the characterisation of spatial rainfall for hydrological modelling. An Islamic Development Bank-funded research student investigated rainfall-runoff performance for selected catchments in the Sultanate of Oman, demonstrating the limited utility of conventional models in catchment analysis, and the relative power of simpler data analysis. Other recent research with PhD students has analysed the spatial modelling of rainfall and snow precipitation using remote sensing and ground-based data for hydrological modelling in Iran, the effect of climate variability and climate change on water resources in Botswana, and the effects of climate variability and land use change on the Kyoga catchment in the upper Nile, Uganda.

Since arriving in Canada, Dr Wheater has developed a new large scale model of Prairie hydrology, and is working on the development and analysis of water resources simulation models for the Prairie Provinces.

In the last 6 years, earlier work on water quality modelling has been published, including concerning the development and UK application of catchment scale phosphorus models, including limitations of data support and associated issues of model identifiability, and the development of a general water quality modelling tool, and its application to rivers in the United States and China. More recent work focussed on the development of new models of nitrates in groundwater-dominated catchments in the UK, incorporating the effects of unsaturated zone storage and transport, as discussed above.

At the request of UK Nirex, the UK body formerly charged with responsibility for deep disposal of nuclear waste, Professor Wheater established a long-term research programme into the representation of biosphere processes for the safety assessment of the deep disposal of nuclear waste, focussing specifically on the transfer of radionuclides from contaminated groundwater to soils and vegetation. The objective was to produce credible models to inform safety assessment, and develop associated parameters for a range of soils, vegetation, and key radionuclides of interest. A major field lysimeter experiment was run from 1988-1996, followed by controlled environment soil column experiments from 1997-2004. Partial funding was also provided by the French governmental. Research from 2005-2007 addressed specific modelling issues concerned with the redox-mediated transport of radionuclides, such as radio-iodine and sodium-22, across a capillary fringe, the upscaling of column-scale experiments for field-scale application, and the development of a new strategy for the safety assessment of deep disposal of nuclear waste. A book was published in 2007 on the 19 years of research at Imperial College.