Student’s Name


Professor’s Name


City (State)


Declaration Statement

I hereby certify that all the material contained within this paper is my own work except where it is clearly referenced to others.


Water shortage, climate change, freshwater, water quality, standardized precipitation index (SP1), reconnaissance drought index (RDI), decision-making, vulnerability.


Water resources in the UK have been highly prone to depletion in the resent years. Extensive research done indicates that the future of the UK with regard to water resources is bound to be affected and may eventually lead to adverse effects. Most of the affected sectors of the environment are the agricultural and demographical domains. A rapid change in the climatic conditions, which is a deviation from the norm has affected these areas as we shall see in the paper. Due to such and more reasons as shall be shown in this paper, we shall undertake to understand the various sectors affected and seek to analyse the same using set methods. Our data collection will highly rely on previous papers done by numerous authors and will show a clear depiction of the state of the area chosen – the Elmley Marshes in the UK. In this paper we shall do analysis and discussions of enumerations based on the set hydrological methods of analysis. Through such methods, we desire that this paper will be invaluable for education and insight.

Table of Contents

Declaration Statement. 2

Abstract. 3

Table of Contents. 4


1.1 Statement of the Problem.. 6

1.2 Theoretical or Conceptual Framework. 6

1.3 Statement of the Purpose. 7

1.4 Research Questions. 7

1.5 Significance of the Study. 7

1.6 Definition of Terms. 7



3.1 Research Questions. 23

3.2 Study Design. 24

3.3 Study Context and Intervention. 25

3.4 Data Collection. 26

3.5 Data Analysis. 29

3.5.1 Standardized precipitation index (SPI). 33

3.5.2 Standardized precipitation–evapotranspiration index (SPEI). 34

3.6 Ethical Considerations. 36

3.7 Assumptions, Delimitations, and Limitations. 36

3.7.1 Assumptions. 36

3.7.2 Delimitations. 36

3.7.3 Limitations. 36



Suggestions for Future Work. 44

Reference List. 45


The conceptual framework surrounding UK water resources with regard to climate change is indicative of the fact that further environmental degradation will result in the deletion of these resources rendering the UK into drought and despair. O’Hara (2007, p. 23) further asserts that with the presently articulated consistency of climatic changes, water systems are bound to be adversely affected even on a global scale. At the forefront of such a worrying host of problems, is an increase in world temperatures. Hydrological impacts as illustrated by Thompson, Gavin, Refsgaard, Sørenson and Gowing (2009, p. 1) – solely based on water quality and quantity – have an increased effect on water resources in the world today. As the world progresses into an age of constant growth and development, there is a risk of losing environmental consciousness since people are more inclined to self-gratification rather than conservation of the resources that sustain us.

One of the best ways by which we can mitigate the effects of and curtail the extensive degradation of the environment and as a result, the water resources, according to, Yousefpour, Temperli, Jacobsen, Thorsen, Meilby, Lexer, Lindner, Bugmann, Borges, Palma, and Ray, (2017, p. 2), is by implementation of management activities. With the candid realization that forests are the major water catchment areas, it is imperative that measures are taken to protect their depletion through deforestation and human encroachment. Hattermann, Post, Conradt, and Wechsung, (2008, p. 42) further assert that, increases climatic changes will eventually have an effect on the agricultural sector. Consequently, we shall be unable to produce enough food to sustain us and therefore human life will be prone to absolute termination and extinction.

To assert the current state of the situation, we shall specifically focus on the effects and causes of climatic change leading up to the adverse effects on water resources in the UK. According to Watts, Battarbee, Bloomfield, Crossman, Daccache, Durance, Elliott, Garner, Hannaford, Hannah, and Hess (2015, p. 7), in the last 50 years, there has been an increase in the intensity of winter rainfall. Their paper further claims that a continuation of such rainfall intensity in the UK shall eventually affect the water ecosystem, the availability and the quality of water in the nation. In this paper we shall assess the current state of water resources in the UK, the ongoing changes in climate conditions, the effects of the same and some of the means used to mitigate the specific adverse effects. We shall undertake several methods of data collection and analysis to realize accurate and reliable results for study.

1.1 Statement of the Problem

It is possible that within the next 100 years, the UK will be prone to increased water shortage. With the influx in industrial development, technology and increased requirement for space for development to cater for the surging population in the world, our resources are at a risk of strain and eventual depletion. Despite the fact that there is increased human activity in the world today, there is little or no concern for the environment. For instance, the paper by Zhou, Q., Mikkelsen, P.S., Halsnæs, K. and Arnbjerg-Nielsen, K., 2012, p. 542) claims that, in the cost-benefit analysis (CBA) – a tool used to analyse economic scenarios where climate change is involved – the social discount rate in nations like the UK has been reduced to 3.5% from 6%, thereby indicating that the environment has been left at a precarious position of probable and eventual destruction. Through this paper, we hope to critically analyse the water resources in the UK and assert whether the present research conducted is in tandem with what we realize and as a result conduct measures to alleviate adverse consequences.

1.2 Theoretical or Conceptual Framework

The sole purpose of this study is to assert whether the UK has undergone extreme fluctuation of the key variables associated with climatic changes, in essence, flood and drought, for example: temperature, soil moisture and rainfall among others. With the in-depth study of these factors, our study aims to adopt the invaluable use of various methods of prediction using indices like the Standardised precipitation index (SPI) and the Reconnaissance drought index (RDI). Using such powerful mathematical and analytical tools, the research aims to provide detailed and specific investigational evidence.

1.3 Statement of the Purpose

The sole purpose of this research is to garner attention to the rapidly changing climate and provide candid proof of such claims using the above stipulated methods. With such tools of research and analysis we shall analyse past, present and future trends and eventually come up with a detailed research paper for the same, within the constraints of the UK’s geographical area.

1.4 Research Questions

Some of the questions we hope to answer through the undertaking and conclusion of this research are:

  1. What is the water storage capacity in the UK and how much water is consumed?
  2. What is the present climatic state of the UK and what are its differences in the past?
  3. What are the various methods used to predict climatic conditions by researchers?
  4. Which of the above asserted models shall we use for this research?
  5. What do the study results acquired from the research show?
  6. Based on the analysis and results acquired from the research project, are there any suggestions for the future?

1.5 Significance of the Study

The study in this research project is important in in the realization of trends in climatic change in the UK. Considering the recent changes in various weather aspects such as rainfall intensity and rise in temperatures, it is essential that we critically collect and analyse data that will lead us up to the realization of the crux of the matter as far as climate change is concerned.

1.6 Definition of Terms

Some of the terms used in this research project are:

Standardised precipitation index (SPI) – a widely used index to characterize meteorological drought on a range of timescales. … It quantifies observed precipitation as a standardized departure from a selected probability distribution function that models the raw precipitation data.

Reconnaissance drought index (RDI) – it is one of the widely used indices, due to its high sensitivity and resilience. The basic form of the index is the ratio of the cumulative precipitation to potential evapotranspiration, for a specified reference period.

Standardized precipitation–evapotranspiration index (SPEI) – is an extension of the widely used Standardized Precipitation Index (SPI). The SPEI is designed to take into account both precipitation and potential evapotranspiration (PET) in determining drought.

Hydrology – A branch of science concerned with the properties of the earth’s water, and especially its movement in relation to land.

Hydrological methods/models – the characterization of real hydrologic features and system by the use of small-scale physical models, mathematical analogues, and computer simulations

Pluvial flooding – A pluvial flood occurs when an extreme rainfall event creates a flood independent of an overflowing water body.

MIKE 11 – A one-dimensional hydraulic modelling system able to represent hydraulic structures including weirs, gates and culverts. The dynamic coupling of MIKE SHE and MIKE 11 evaluates for each time step river-aquifer exchange, overland flow to channels and flooding from channels to adjacent grid squares.


In the recent past, there has been an up soar of increased climatic changes which is exhibited by erratic rainfall patterns more so in the urban areas. As such, Olsen, Zhou, Linde and Arnbjerg-Nielsen (2015, p. 256) claims that such weather changes may result in pluvial flooding therefore affecting urban settings to great extents. The drainage systems in urban areas have been built based on historical weather and climate patterns which have changed sporadically over time. It is imperative that these systems are reserviced or completely changed to accommodate the increasing amounts of drainage water being release by the ecosystem through climate change. Zhou, Mikkelsen, Halsnæs and Arnbjerg-Nielsen, 2012, p. 539-540) claim that to quantify the best way to adapt to the changes in the climate, it would be necessary to increased technological advancements to cater for detailed researches on the risk of pluvial flooding in cities and use this same information to take mitigation measures geared towards improvement of our drainage systems, (Muzik 2001, p. 236). They further assert that since climate change is ongoing even today, we should use the past and present data to predict future events and use that for protectionist measures. Quintessentially, the study and research of climate change with regard to water resources in the UK is important for present and future palpable action.

From a regional perspective, Watts, Battarbee, Bloomfield, Crossman, Daccache, Durance, Elliott, Garner, Hannaford, Hannah, and Hess (2015, p. 7-8) ascertains that there is less clarity on anthropogenic changes in climate. There is an increased variability of precipitation levels in some areas and during some seasons. In UK where the North Atlantic Oscillation, storm track and blocking greatly influence the weather patterns, regional scale outputs need to be realized in order to reflect the actual state, (Watts, Battarbee, Bloomfield, Crossman, Daccache, Durance, Elliott, Garner, Hannaford, Hannah, and Hess 2015, p. 7-8). Without proper conceptual frameworks and accurate deductions from weather and climate changes, Kelman and Gaillard (2010, p. 23), predict that there will be increased vulnerabilities and higher risk for disaster in the UK. With such deductions, we realise the need to undertake this study for future predictions, risk assessments of hazardous areas and viable mitigation factors, (Zhou, Mikkelsen, Halsnæs and Arnbjerg-Nielsen, 2012, p. 540).

Fig. 1. Framework to assess economic pluvial flood risk
Source: Olsen, Zhou, Linde and Arnbjerg-Nielsen 2015, p. 257
Fig. 2. Flood hazard assessment showing both the extent and the depth of the flooding for a given return period.
Source: Olsen, Zhou, Linde and Arnbjerg-Nielsen 2015, p. 258
Fig. 3. The schematic view of dependencies and relationships within the groundwater-agriculture-climate change triangle. Direction and strength of the dependencies differ under different climatic and socio-economic conditions.
Source: Barthel, Reichenau, Krimly, Dabbert, Schneider and Mauser 2012, p. 1930

According to Barthel, Reichenau, Krimly, Dabbert, Schneider and Mauser (2012, p. 1930-1931), climate change, groundwater and agriculture are highly correlated in complex ways. Such interdependence is the source of direct proportionality of every action and reactional changes in any of these aspects. Pingale, Jat and Khare, (2014, p. 176) claim that based on the uncertainty of the state of vital ecosystems given the tumultuous nature of changes occurring in the climate, there needs to be an equitable management of resources in processes that promote conservationist measures to these water resources. In the process of studying climate change, the use of hydrological regime in this continuous assertion will help in determining the effects of weather and climate change on the agricultural state of the land, (Hattermann, Post, Conradt and Wechsung 2008, p. 42). In furtherance thereof of this study, eco-hydrological methods could be used in climatic change assessments to realize crop yield quantities in the soil and water integrating model (SWIM), (Hattermann, Post, Conradt and Wechsung 2008, p. 42).

Fig. 4. Different order stages of climate change effects on water resources
Source: Gain, Giupponi and Renaud 2012, p. 346

As illustrated in the diagram above, Gain, Giupponi and Renaud (2012, p. 346-347) illustrate that there are various levels to which climate change and its continuity are able to continually affect the water resources while destroying cities and civilization. O’Hara (2007, p. 23) further claims that it is anticipated that climate change and rising temperatures will lead to a flurry of problems. These may include: flooding, melting slow, increased surface runoff, inadequate drainage systems to handle increased drainage water levels and precipitation frequencies thereby reducing the amount of fresh water that humans require in abundance. In the diagram shown, we are made privy to a possible sequence of events that could occur with the increasing changes in climatic conditions. These ordered stages illustrate that despite the increased development and growth of industrialization of world economies, ecological imbalance resulting from such activities will bring a halt to these activities and plunge the world into crisis. In targeting the UK as our geographical location for study, we place ourselves in a position to gather extensive data and recurring events that will help formulate our thesis and come up with a conclusive inference. The eventual conclusion of such events is a backward and damaged society in problem and suffering, (Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 503-504). Through this research, we hope to alleviate such consequences by offering suggestions for implementation with reference to the results obtained.

Olmstead (2014, p. 2), in his study, illustrates that water resource adaptation methods need to be developed in order to cater for the world’s consumption levels. His study further affirms that irrigation accounts for 70% of the globe’s water withdrawal and 90% of human consumption. These staggering numbers indicate that the world is heavily reliant on water resources and a depletion of the same will realise adverse effects. Scientific research has also shown that 70% of the human body is made up of water. Such facts further assert the reasons as to why this specific study on the nature of water resources in the UK in light of climate changes, is necessary. Considering the unsustainability of unreliability of vulnerability measures since its assessment is not clear, Gain, Giupponi and Renaud (2012, p. 346), management strategies for our limited water resources should be enhanced for implementation of sustainable methods of water resource usage.

Fig. 5. An illustration of climate change expectations seen at time point ‘now’ i.e., Enow (Climate) to the several decision points in the future
Source: Yousefpour, Temperli, Jacobsen, Thorsen, Meilby, Lexer, Lindner, Bugmann, Borges, Palma, and Ray, (2017, p. 5)

Davies and Simonovic (2011, p. 684-685) claim that there is a constantly increasingly demand for fresh water in the UK and the world as a whole. Despite this increasing need, pollution, industrialization, population growth and climate change are some of the numerous factors that constrain the achievement of this need. Since we cannot control the whims of nature and the ecosystem, it is upon us to take part in conservationist measures that protect and offer sustainability to the environment. Being the greatest global crisis that humanity has ever faced, as Kelman and Gaillard 2010, p. 30) assert that, there will be a further and catalysed destruction of water and forest resources and therefore we should utilize them when they are available. The above notion is majorly harboured by some researchers who have realized that the present climate changes are continually in furtherance of world destruction and that despite whatever measures that persons and individuals put to try rectifying the situation, the damage has been done already. The despair harboured by such a notion advocates for overexploitation of what we presently have and enjoy thereby assuming that we have no part in ensuring that future generations enjoy the same resources that we do presently. As such, this research hereunder and analytical methods are aimed at realizing the extents to which we have destroyed the water resources in the UK and the measures we could take to ensure that we change this overall and eventual doom that awaits us if changes are not promptly made.

While an overview of the various available methods of study is essential for research and understanding, this specific research project will categorically base its analysis and study on the hydrological model. Through this method, we shall obtain data and in turn, valuable information on the processes involved in climatic change and its effects through and on urban developments, (Praskievicz and Chang 2009, p. 650). Consequently, we shall understand potential ranges of these effects on water resources in the present and for the future. To better understand the study, we need to initially understand the terminologies used. For instance, hydrology, which can be defines as the study of water movement in the earth with respect to the land mass. In the same vein, the understanding of hydrological modelling which involves the prediction and management of earth’s water resources is also necessary. Murray, Foster and Prentice 2012, p. 14) claim that since biospheric processes are essential in hydrological cycle regulation globally, then we should involve them in simulations on water resource research. With such an extensive scope of research in mind, we shall dare to narrow down to the intricate details that will help us realize the statement of problem and viable solutions for development and longevity of the human race through preplanning processes.

Climate change in the UK has caused various extensive consequences on water resources, the economy, population growth and distribution, technology, social and economic factors, (Alcamo, Flörke and Märker 2007, p. 247-248). These changes not only affect human life, but also plant and animal life. With increased sedimentation in rivers and lakes due to surface runoff, there are increased amounts of chemicals entering water bodies and affecting aquatic life. With such events occurring, the blue economy in the UK is greatly affected yet it is a major source of income to the government. The build-up of such events causes a ripple effect where people lose jobs, sources of livelihood and even places of habitation. Consequently, the water resources are depleted and the lives of humanity destroyed. Praskievicz and Chang 2009, p. 651) claim that, to foster understanding in possible future conditions resulting from climate change, then hydrologists heavily rely on available models to conduct simulations for potential future behaviours and patterns. Such simulations offer a basis of argument on how the water resources have been utilized and affected presently, to how they are bound to change and become in the future if the current trends in utilization are continued. These models however cannot be accurate in stand-alone form and therefore they are used in tandem with other research theories to find solutions to the problem of climatic changes in the UK and the world.

Flato, Marotzke, Abiodun, Braconnot, Chou, Collins, Cox, Driouech, Emori Eyring and Forest (2014, p. 743-746) claim that there is a wide range of models that can be used for climate research ad these range from simple balanced energy models to the extensively complex earth system models (ESMs). Therefore, with the expanse in choices for models, it is only possible to choose one model depending on the complexity of the parameters to be measured and time frames to be used. For instance, some of the most prevalent applications for these climate research models include: historical climate sensitivity, the simulation of paleo and prediction of variability in near-term climate. With the extensive nature of options however, comes the need to use an option that maximizes on the costs of computation, (Flato, Marotzke, Abiodun, Braconnot, Chou, Collins, Cox, Driouech, Emori Eyring and Forest 2014, p. 746). As per the arguments made by Randall, Wood, Bony, Colman, Fichefet, Fyfe, Kattsov, Pitman, Shukla, Srinivasan and Stouffer (2007, p. 618) hydrological research methods for climate changes may not be most efficient in some aspects but could be beneficial in the prediction of streamflow; resultant from solar irradiance and the composition of the atmosphere. Therefore, this implies that hydrological methods, as shall be discussed herein are to some extent beneficial to the study of climate change and the prediction thereof, but cannot, in a real case, be used independently for research.


In the study of water resources in the UK, it has been realized that drought has been the most probable eventual effect of the rising temperatures and erratic precipitation. As such, we realize the need to use research methods that prioritize on the highest risk factor which is drought, (Salimi, Asadi and Darbandi 2021, p. 1-2). The authors further state that hydrological drought could be construed to occur after a meteorological drought, hence the need to monitor such cases using drought indices such as the standardized precipitation index (SPI) – majorly used in monitoring meteorological droughts which naturally occur faster. Pathak. and Dodamani 2019, p. 1) also claim that these available indices can only be effective if employed in the correct climatic regions. The above is because, indices like the SPI are less sensitive to low rainfall and therefore might underestimate results from its extremities. The drawback by SPI resulted in the development of RDI and SPEI which were mitigation factors for the underestimation done by SPI. Since different regions have diversified flora, fauna, humidity, temperature and land use, we cannot possibly ascertain drought indices for the world. We can however use regional data to derive conclusions that are accurate and actionable enough, hence our choice to study water resources in the UK.

Fig. 6. A schematic of water balance model and interactions
Source: O’Hara 2007, p. 30

The Water Balance Model is a series of equations that simulate the conservation of water volume for a regulated watershed on a monthly time scale. A significant modification to the abcd model that is necessary for our methodology is the inclusion of imports into the system of water balance equations. With such modification the model represents a generally applicable tool for the purposes of climate change adaptation studies in urban environments in semi-arid or arid regions. Many cities in semi-arid regions, in the UK, rely on local runoff to meet up to 20% of water demand and imports the remainder of the demand amount. For this case study, the abcd model component that simulates groundwater withdrawals is not used as no such withdrawals exist. Figure 6 provides a schematic overview of the components of the water balance model as applied to the case study and their interconnections. Precipitation and temperature are the main climatic drivers of the water balance model.

Precipitation estimates serve as input to the local runoff model component of the water balance model for the estimation of expected urban water consumption targets. Temperature estimates are used for the determination of water consumption and reservoir evaporation. Urban population estimates are also used for the estimation of expected imports and water consumption. Water in reservoir storage is replenished by local runoff and imports. It is depleted by evaporation from the water surface in the reservoirs, releases to meet expected water consumption and any losses due to enforced spillage to avoid overtopping. The model mathematical formulation is shown next in discrete form for a typical month t. To keep variable symbols to a minimum, the equations are generalized to show dependence on the kth member of the uncertain variables.


In the above equations:

  • t is the time index for a monthly time interval of computations,
  • k is the ensemble-member index,
  • j is a season index to indicate parameter dependence on season, and
  • i is a monthly index to indicate parameter dependence on a particular month.
  • Q, P, I, and E denote monthly volumetric flow quantities: reservoir release, basin precipitation and total reservoir water surface evaporation.
  • denotes the water volume in reservoir system storage at the beginning of month t and for ensemble member k. The symbol,  represents the stock of water at the end of month t and for ensemble member k.

In this formulation, spillage capacity is taken to be a function of the actual reservoir capacity – denoted by K – times a multiplier .

Assuming that monthly precipitation is distributed uniformly over the natural drainage basin and that runoff is a linear function of precipitation for monthly aggregate quantities, for each realization of precipitation we use:

The expression is implicit where  is a parameter depending on season j. It implies that only a fraction of precipitation enters the reservoir each period, and that the rest is lost to evapotranspiration or percolation to deep groundwater storage in the upstream drainage basin. The parameters di,  and  are calibrated to fit historical data.

In the hydrologic model parameter estimation, parameter b depends on the season with two parameter values defined – one for the months in winter and the other for the months in summer. We find these parameters by comparing actual precipitation with actual runoff data. We use the random sampling method for estimation of parameters. We impose a uniform probability distribution U (0,1) upon the possible range of values for b. Using a random number generator, we choose the pair of b values that lead to the smallest mean square error when compared to the runoff data. The release parameter d takes an average value when available reservoir rule curves are used. We make monthly adjustments to the value of d because releases occur at different points in different months. We calculate the release fraction for a given month of the year by multiplying d by the ratio of the release for that month to the average release for all months of the year. Spills occur as a result of reservoir inflow while reservoir content is at capacity. We define the parameter α as the highest percentage of capacity that the system can hold without spills occurring. For model simulations, αK denotes the capacity constraint for which the spills occur. In the absence of daily data, reasonable approaches to estimation of a value for α from the historical record is to use:

o   t denotes a month for which historical spills occurred, and
o   TN is the total number of months with spills

We obtain a second estimate for the value of α that preserves the historically observed total spill volume and use this new estimate in the sensitivity analysis with respect to spills. The new estimate of α, denoted as α’, yields a higher spill volume in accordance with historical records.

Fig.7. Isle of Sheppey and the Elmley Marshes, North Kent, UK
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505

The specific study site chosen was the Elmley Marshes located on the southern side of the Isle of Sheppey at the end of the Thames Estuary, southeast England. They are part of the wider North Kent Marshes, the largest tract of coastal wet grassland remaining in England and Wales. They are typical of lowland wet grassland, a wetland type which includes semi-natural floodplain grasslands, grazing marshes, flood meadows, manmade washlands and water meadows as illustrated by Jefferson and Grice (1998); Joyce and Wade (1998). Wet grasslands are characterised by periodic, but not continuous inundation, and permanently high-water tables. They are predominantly located in river valleys, areas of impeded drainage or behind sea defences. The North Kent Marshes were created by the progressive enclosure and drainage of salt marshes. Evidence of these activities remains in the form of embankments which delineate the boundaries of earlier enclosures. Old embankments define the northern and eastern boundaries of the Elmley Marshes, whilst to the south and west they are bounded by the current sea defences. The embankments therefore effectively impound a discrete hydrological unit of around 8.7 km2. Two small hills in the southeast rise to around 12 m above the marsh surface which itself has a mean elevation of 1.90 m above Ordnance Datum (m OD).

Fig. 8. A topographic grid and model of the Elmley Marshes
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505

There is a gradual decline in elevation towards the south where mean elevation is 1.75 m OD. Soils are of the Wallasea– Downholland association and comprise peloalluvial gley soils derived from non-calcareous, clayey marine alluvium as Fordham and Green (1980) assert. High clay content results in low permeability and slow rates of water movement (Hazelden et al. 1986). The drainage system comprises a ditch network dividing the marshes into fields. It reflects the original salt marsh drainage, although ditches have been straightened, widened and deepened to improve drainage efficiency. Five main ditches cross the site into which secondary ditches converge. Additional drainage features are shallow linear features (rills) superimposed on field surfaces. These are remnants of small-scale salt marsh drainage channels. Tidal outfalls at the downstream ends of the main ditches discharge water at low tide into the Swale, a tidal channel separating the Isle of Sheppey from the mainland. Immediately upstream of these outfalls and at four other locations water control structures are installed within the ditches (Fig. 2). These are ‘‘drop board sluices’’ comprising a grooved concrete spillway into which wooden boards are inserted or removed to control water levels (Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, pp. 503-523). Impoundment of the Elmley Marshes within embankments, coupled with low hydraulic conductivity, restricts surface and groundwater inflow so that seasonal differences in the relative importance of precipitation and evapotranspiration drive the hydrological regime.

Preferential flow through macropores formed by summer desiccation of clay soils promotes rapid autumn rise in water tables. Low-lying areas, including rills, are initially saturated and produce the first runoff to the ditches. Ditch water levels rise in response to this runoff and direct precipitation until water spills over the drop board sluices. At this time rills may become connected to ditches and act as pathways for water movement towards field centres (Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, pp. 503-523). The water table is at, or close to, the surface for much of the winter and early spring creating a mosaic of dry land and shallow flooding which is supplemented by ponding of precipitation and runoff. Through spring and summer, the water table and ditch water levels decline. Rills and other microtopographic depressions are the last areas to hold surface water. By mid-summer the water table reaches 0.80– 0.95 m below the surface. Ditch water levels experience a similar decline and the shallowest may dry out. Shallow water tables and winter flooding limited the traditional use of wet grasslands to extensive or low intensity agricultural activities associated with grazing and hay cutting. Grazing by sheep and cattle throughout the whole of the Elmley Marshes produces short to medium-length grass swards.

Adas (1997, p. 82-89) classed the dominant vegetation communities according to the National Vegetation Communities system, (Rodwell 2006, p. 1-66) as MG6 (Lolium perenne/Cynosurus cristatus grassland), MG7 (Lolium perenne leys) and MG11 (Festuca rubra/Agrostis stolonifera/Potentilla anserina grassland). Specialist grassland species indicative of wet conditions within coastal grazing marshes include sea arrow grass (Triglochin maritima), divided sedge (Carex divisa) and saltmarsh rush (Juncus gerardi). Ditch flora reflects reductions in salinity inland from the tidal outfalls (Hollis et al. 1993; Hollis and Thompson 1998). Brackish communities with species including sea club-rush (Scirpus maritimus), fennel pondweed (Potamogen pectinatus), and soft hornwort (Ceratophyllum submersum) are found in most ditches, whilst fresher areas inland have more diverse flora. Species here include lesser reedmace (Typha angustifolia), glaucous bulrush (Schoenoplectus tabernaemontani) and fool’s water cress (Apium nodiflorum). Ditches also support a good invertebrate community. Beetles, dragonflies and damselflies are particularly well represented (English Nature 1990). Low-intensity management coupled with high water tables and flooding have contributed to the importance of wet grasslands as wildlife habitat especially for waterfowl and wading birds (Ausden et al. 2001; Ausden and Hirons 2002).

The Elmley Marshes support 70% of the breeding waders of North Kent including nationally important breeding populations of lapwing (Vanellus vanellus) and redshank (Tringa totanus), as well as other notable species such as black tailed godwit (Limosa limosa). They are listed as internationally important under the Ramsar Convention, are included within a special protection area under the EC Directive on the Conservation of Wild Birds (79/409/EEC) and include a number of sites of special scientific interest. The Elmley Marshes are a National Nature Reserve and are managed in line with agri-environmental schemes designed to promote ecological friendly farming including the maintenance of ecologically driven water levels and restoration of wet grassland in areas converted to arable. These measures aim to redress the loss of wet grassland, largely due to agricultural intensification and associated drainage and flood defence works, which took place during the second half of the twentieth century (Williams et al. 1983; Williams and Hall 1987; Mountford 1994). These losses are implicated in the decline in wetland-related species including lapwing and redshank (Green and Robins 1993; Ausden et al. 2001).

3.1 Research Questions

  1. What is the precipitation range of the area in the past and the present?
  2. What are the effects from the same?
  3. How has climate change affected the numbers?
  4. What do we predict for the future?
  5. Is the research indicative of future drought risks?
  6. How are water resources getting affected and have we felt such effects presently?
  7. How does society have a role to play in curtailing adverse effects of climate change?

3.2 Study Design

Our model uses a finite difference approach to solve the partial differential equations describing overland (two-dimensional Saint-Venant equation), unsaturated (one-dimensional Richards’ equation) and saturated subsurface flows (three-dimensional Boussinesq equation). Analytical solutions are used for describing interception, evapotranspiration and snow melt. Channel flow is simulated using MIKE 11, a one-dimensional hydraulic modelling system which can represent hydraulic structures including weirs, gates and culverts (Havnø et al. 1995). The dynamic coupling of MIKE SHE and MIKE 11 evaluates for each time step river-aquifer exchange, overland flow to channels and flooding from channels to adjacent grid squares. Thompson et al. (2004) provided a detailed account of the development, calibration, validation and results of a MIKE SHE/MIKE 11 model of the Elmley Marshes, one of the first applications of the coupled modelling system to a wetland. Data employed within the model are summarised in Table 1. The model area was divided into 9,271 grid-squares of 30 m 9 30 m with the elevation of each provided by a 1:2500 topographic map (Fig. 2). A single uniform saturated ozone layer was specified and its hydraulic conductivity varied during calibration from an initial value guided by Al-Khudhairy et al. (1999) and Gavin (2001). A zero-flow boundary around the model area was specified due to the impoundment of the marshes within embankments and the low hydraulic conductivity. The drainage option was used to represent runoff from topographic features too small to be shown in the model grid. A uniform soil profile with hydraulic properties based on Al-Khudhairy et al. (1999) was specified and included bypass flow to represent macropores. Precipitation and Penman– Monteith potential evapotranspiration (Monteith 1965) were provided by an automatic weather station. These data were supplemented by a nearby rain gauge and the UK Meteorological office rainfall and evaporation calculation system (MORECS, Meteorological Office 1992). Evapotranspiration parameters for a uniform grass cover were taken from the literature (Table 1). Overland flow resistance was a calibration term with initial values taken from Al-Khudhairy et al. (1999). The MIKE 11 ditch network was abstracted from 1:2500 digital map data (Ordnance Survey Landline Plus). Cross-sections were based on field surveys, aerial photography and literature (Newbold et al. 1989). Uniform channel roughness and leakage coefficients were used and were both modified during calibration. Rectangular weirs, with dimensions from LMIDB (1999), represented drop board sluices. The specification of positive flow only valves on the weirs at the downstream ends of the main ditches ensured they operated as tidal outfalls. Evaporation from ditches was represented as boundary conditions at the end of reaches which abstracted volumes of water derived from the product of daily evaporation rate and ditch water surface area. The latter were evaluated from water level/surface area relationships and water level records from stage boards and an automatic water level recorder (AWLR). The MIKE 11 model, including the location of cross sections, weirs and evaporation boundary conditions, is shown in Fig. 2. A 36-month simulation period (25/06/1997–29/06/ 2000) was divided into two 18-month sections for split sample calibration and validation (e.g., Klemesˇ 1986; Refsgaard 1997). This was based upon graphical comparisons of observed and simulated water table depths (obtained from piezometers) and ditch water levels from stage boards and the AWLR. Nash– Sutcliffe efficiency coefficients (R2, Nash and Sutcliffe 1970) were also evaluated. Good agreement was obtained between model results and observed hydrological conditions. For example, Figs. 4 and 6 provide representative observed and calibrated groundwater depths and ditch water levels (in addition to the result of climate change scenarios discussed below). The R2 values for these comparisons are 0.80 and 0.83, respectively.

3.3 Study Context and Intervention

This study employed the climate impact assessment methodology advocated by Parry and Carter (1998) which has been widely used to assess hydrological impacts on river discharge (e.g., Chiew et al. 1995; Viney and Sivapalan 1996; Limbrick et al. 2000; Menzel and Bu¨rger 2002). Arnell and Reynard (1996) outlined the stages that this approach involves: 1. Define, calibrate and validate a hydrological model using current climate data; 2. Define climate change scenarios and perturb the original model input data accordingly;

3. Run the hydrological model using new input data and compare results with those obtained for current climate conditions. The development, calibration and validation of the hydrological/hydraulic model of the Elmley Marshes by Thompson et al. (2004) satisfy the first of these stages. Results of this model provide the baseline conditions against which the impacts of climate change can be compared.

This study uses climate changes predicted for the 2050s by the UK Climate Impacts Programme (UKCIP02, Hulme et al. 2002). They are based on mean climate during a time slice covering 2041–2070 driven by four greenhouse gas emission scenarios described by the IPCC Special report on emissions scenarios (SRES) (IPCC 2000). Changes in climate parameters for the Low Emissions, Medium–Low Emissions, Medium–High Emissions and High Emissions scenarios (referred to here as L, ML, MH and H, respectively) are referenced to a 1961–1990 baseline period. These changes are derived from a nested climate modelling approach in which a global climate model (HadCM3, resolution 250–300 km) provides boundary conditions for a global atmospheric model (HadAM3H, *120 km) which in turn provides boundary conditions for a regional model of the European atmosphere (HadRM3, *50 km). This dynamic downscaling provides a more appropriate resolution of climatic outputs from global climate models for use in hydrological impact studies (e.g., Kay et al. 2006; Fowler and Kilsby 2007). Figure 3 summarises predicted changes in four climate parameters, expressed as average departures from the baseline period, for the Elmley Marshes for each emissions scenario for the 2050s time slice. The southeast of the UK, already the driest part of the country, is subject to some of the largest changes projected by the UKCIP02 scenarios. Figure 3a shows that summer (August) temperatures in the 2050s are projected to be 2.1C warmer under the Low Emissions scenario and 3.3C higher under the High Emissions scenario. Even in winter (January) temperatures are between 1.2 and 1.9C warmer. In winter precipitation increases whilst summers are drier (Fig. 3b). These changes range from a 10.3–16.4% increase in winter (January) to a 19.9–31.6% decline in summer (July). Due to reduced cloud cover in every month except January downward shortwave flux, which comprises both direct and diffuse solar radiation, is higher (Fig. 3c). This will further increase elevated evaporation rates resulting from higher temperatures. Changes in wind speed will also impact evaporation and Fig. 3d shows that wind speed is predicted to increase by 2.8–4.5% in winter and decrease by 1.9–2.9% in summer. The degree of confidence in these projections is, however, lower than for some other parameters including temperature and precipitation (Hulme et al. 2002).

3.4 Data Collection

Table 1. Annual precipitation, potential evapotranspiration and net precipitation for the Elmley Marshes in 1997/1998 and 1998/1999
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505
Fig. 9. Average departures from the baseline period for the UKCIP02 2050’s time slice and the four emissions scenarios: a. Temperature (C), b. precipitation (%), c. downward short-wave flux (W m2), d. wind speed (%)
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505
Fig. 10. Simulated groundwater depths within the Elmley Marshes for the calibrated model and climate change scenarios: a. Using PETtemp, b. using PETtrws. Observed groundwater depths used in model calibration are also shown (R2 = 0.80)
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505
Table 2. Number of days when simulated water tables are above threshold depths close to the ground surface during each complete hydrological year of the simulation period for each scenario
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505
Table 3.
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 505

3.5 Data Analysis

Changes in the hydrological regime of the Elmley Marshes revealed by the simulation of climate change scenarios are summarised. Climate change has the effect of tipping the balance between precipitation and evaporation in favour of evaporation thereby inducing drier conditions. The autumn/winter rise in water table and ditch water levels are delayed whilst in the spring levels begin to decline earlier. Periods of high-water levels are therefore shorter. Consequently, the extent and duration of extensive surface inundation both declines. Spring and summer recessions in groundwater and ditch water levels are enhanced and continue for longer into the following autumn. The magnitude of these changes generally increases through the progressively higher emissions scenarios and are particularly pronounced when predicted changes in temperature, net radiation and wind speed are used to modify evapotranspiration. In these cases, winter groundwater levels rarely approach the ground surface, limiting inundation. Flooding is further impacted for these PETtrws scenarios as ditch water levels do not reach the threshold levels required to induce movement of water from the ditches onto the marsh surface. Where enhanced winter precipitation associated with predicted climate change raises water levels, these changes are modest, especially when compared to the magnitude of reductions in water levels at other times, and short-lived. The impacts of climate change upon the hydrology of the Elmley Marshes are likely to have a number of ecological implications. Since it is not unreasonable to suggest that other parts of the North Kent Marshes and similar wetlands in southeast England can be expected to experience broadly similar changes in their water level regimes, these ecological impacts are of relevance beyond the Elmley Marshes.

Wetland plant communities and species have specific and critical ecohydrological requirements which include water level regime (Wheeler and Shaw 1995; Wheeler et al. 2004). Wheeler et al. (2004) advocated the combination of water level requirements of wetland plant communities or species and predicted hydrological changes from modelling to evaluate impacts of climate change or other scenarios involving, for example, abstraction or restoration. They suggested that this approach could establish whether vegetation is ‘out of regime’ or is at risk of moving out of regime in terms of water needs. This approach does, however, require hydrological models capable of accurately simulating water table elevations at a resolution that is commensurate with hydrological changes that would induce ecological impacts. For example, whilst a model might be considered accurate if it simulates groundwater level within 10 cm of observed water tables a difference of this magnitude is likely to provide conditions associated with very different vegetation communities (Wheeler et al. 2004). The present study brings these model and ecological scales together thereby enabling an examination of the potential ecological impacts of the climate change scenarios. Application of the sum exceedance approach proposed by Sieben (1965) and adapted to wet grassland communities by Gowing et al. (1998) confirms that the soil water regime of the calibrated model, which represents current conditions, corresponds to floodplain grassland experiencing moderate stress from soil anoxia resulting from high water tables. In the Elmley soils, anoxia is likely to occur when water tables are shallower than 0.3 m as air-filled porosity of the root zone falls below 10% (Taylor 1949). All four PETtemp scenarios result in stress from anoxia being reduced by approximately half while the PETtrws scenarios remove this stress entirely. Absence of soil anoxia will favour grassland communities that are typical of well-drained fertile soil (e.g., MG6) at the expense of those tolerant of waterlogging (e.g., MG11). The latter currently characterises the Elmley Marshes and include specialist species of coastal grazing marshes, such as sea arrow grass (Triglochin maritima), divided sedge (Carex divisa) and saltmarsh rush (Juncus gerardi). These species are likely to be lost from the sward and replaced by more generalist species typical of dry grassland. Ditch vegetation may also be impacted by climate change induced lower water levels. More ditches are likely to dry out in summer, whilst an expansion of brackish conditions into areas that are currently relatively fresh and support more diverse flora may also occur.

Drier conditions are also likely to impact wader populations, which provide the marshes with particular conservation significance. Distributions of wading birds on wet grasslands are strongly related to surface wetness (e.g., Eglington et al. 2008). The probability of a particular part of the marshes being occupied by redshank and lapwing during the breeding season (April–June), as well as wader density, increases with flood extent and the number of wet rills and hollows (Milsom et al. 2000, 2002). Feeding rates are also higher in rills which are wet in May compared to those which are dry (Milsom et al. 2002). This may result from the effects of prolonged inundation on vegetation cover, availability of aquatic invertebrates or more penetrable wet soil. Under calibrated conditions there is extensive shallow inundation in April and May (Figs. 8, 9). The mean extent of inundation of 0.1 m maximum depth during these months for 1998 and 1999 is 1.21 and 1.13 km2, respectively (Fig. 10a, b). Flooding at this time is concentrated in the southern low-lying parts of the marshes and further north alongside ditches. Isolated topographic lows are also flooded (Fig. 10c). For the Ltemp and Htemp scenarios the mean extent of shallow flooding in April and May declines by 25.8 and 48.1%, respectively (Fig. 10a). Relatively large areas in the south and adjacent to ditches are still inundated although away from these locations flood extent declines especially for Htemp (Fig. 10d, e). Maintenance of some inundation suggests that, at least in terms of water requirements, the marshes would be able to support lapwing and redshank although numbers may decline. Much larger reductions in breeding season flooding result from the PETtrws scenarios. The mean reduction compared to calibrated results for 1998 and 1999 are 79.5 and 90.8%, respectively. Flooding at this time is restricted to small low-lying areas adjacent to ditches in the far south of the marshes. The national press has highlighted the plight of lapwing, redshank and other waders due to drought at locations in southeast England including the Elmley Marshes (The Times 2005; BBC 2006). Lapwing and redshank numbers on the marshes plummeted in 2005 due to low spring water levels (Burston 2006).

The PETtrws scenarios suggest a shift away from the hydrological conditions which have favoured wading birds and a decline in the ability of the marshes to support large numbers of lapwing and redshank. These, and other species of national importance such as black-tailed godwit, will face sub-optimal habitat conditions, posing serious threats to their breeding success. Since these birds display philopathy (breeding site loyalty) failure to breed or low chick survival rates will negatively impact on future populations. Similar hydrological changes could be expected throughout the North Kent Marshes so the ability of these species to find alternative sites will also be compromised due to both physical habitat changes and competition with other birds. This is likely to have long-term implications for the wider wader populations. Studies, such as the one presented in this paper, are important in order to determine potential hydrological and resultant habitat changes associated with climate change and to guide management. Opportunities for tackling the hydrological impacts of climate change within the Elmley Marshes are, however, limited. Maintenance of relatively high surface water levels throughout spring and summer will rely on storing as much water as possible in winter to counteract higher evaporation rates. Thompson (2003, 2004) showed that current water level management, which sets the drop board sluices at mean field level, approaches the optimum in terms of maintaining high ditch water levels and inundation. Even under current climate conditions only very modest increases in ditch water level and flooding could be achieved by raising sluices. Mitigating the impacts upon groundwater levels will be more problematical. The low soil hydraulic conductivity limits interactions between ditches and shallow groundwater (Gavin 2001; Thompson 2004). Even if predicted declines in ditch water level could be reduced, there would be little impact on the lower water tables which would result from enhanced evapotranspiration. Flooding the marsh surface supplements the water table but the extent of inundation has been shown to decline progressively under each of the higher emissions scenarios whilst, as noted above, there are limited possibilities for enhancing inundation with existing infrastructure. Limiting discharge through tidal outfalls might offer a partial solution to raising ditch water levels, inducing inundation and supplementing the water table, although this would need careful management. Bunds or low earth embankments have been used in another wetland National Nature Reserve on the Isle of Sheppey (English Nature 1991) and elsewhere in the UK (RSPB et al. 1997) to retain and control water levels within relatively small areas. Lowering small parts of the marsh surface through the creation of scrapes has also been undertaken to prolong surface wetness.

The costs of implementing these approaches for larger areas may, however, be prohibitively large. Similarly, the long-term sustainability of supplementing water levels by pumping from the underlying chalk, an approach adopted elsewhere on the Isle of Sheppey, is doubtful. The marshes face other climate change impacts. Relative sea levels in southeast England are predicted to be between 19 and 79 cm higher by the 2080s compared to 1961–1990 (Hulme et al. 2002). This is likely to limit the period when gravity drainage through tidal outfalls is possible. Silting up of outfalls from an adjacent marsh has already necessitated the installation of pumped drainage (Hollis et al. 1993). The reduction in outflows might, at least initially, act to mitigate impacts of climate change on water levels by facilitating water storage. However, the ability to manage water levels would be limited, and ecologically damaging water levels and flood durations might result. Higher sea water levels will increase the chance of embankments being overtopped and will add to debates over whether sea defences should be maintained to protect what is, at least from an agricultural point of view, less economically valuable land (e.g., Ledoux et al. 2005). The ultimate fate of the marshes may be the replacement of freshwater wetlands with saline ecosystems, a trend forecast for other coastal areas (e.g., Mulrennan and Woodroffe 1998; Eliot et al. 1999).

3.5.1 Standardized precipitation index (SPI)

The standardized precipitation index (SPI) is a widely used index to characterize meteorological drought on a range of timescales. On short timescales, the SPI is closely related to soil moisture, while at longer timescales, the SPI can be related to groundwater and reservoir storage. The SPI can be compared across regions with markedly different climates. It quantifies observed precipitation as a standardized departure from a selected probability distribution function that models the raw precipitation data (McKee et al. 1993). The raw precipitation data are typically fitted to a gamma or a Pearson type III distribution and then transformed to a normal distribution. The SPI values can be interpreted as the number of standard deviations by which the observed anomaly deviates from the long-term mean. The SPI can be calculated (Eq. 1) for differing periods of 1–36 months, using monthly precipitation data. In the operational community, the SPI has been recognized as the standard index that should be available worldwide for quantifying and reporting meteorological drought. Concerns have been raised about the utility of the SPI as a measure of changes in drought associated with climate change, as it does not deal with changes in evapotranspiration. Alternative indices that deal with evapotranspiration have been proposed. Ultimately, the SPI value is calculated by converting the cumulative probabilities based on the Gamma distribution to a standard normal distribution (McKee et al. 1993). The SPI can be calculated as follows:

where x denotes precipitation value; b and c represent the scale and shape parameters of the C function; S is positive and negative coefficients; and c0, c1, c2 and d1, d2, d3 are calculated parameters. Their values are displayed as follows: c0=2.515517, c1=0.802853, c2=0.010328, d1=1.432788, d2=0.189269 and d3=0.001308. G(x) denotes the probability distribution of precipitation. When G(x)>0.5, then H(x)=1-G(x) and S=1; otherwise, H(x)=G(x) and S=− 1.

3.5.2 Standardized precipitation–evapotranspiration index (SPEI)

Penman–Monteith’s method is the standard method for calculating the reference evapotranspiration (Alan et al. 1998):

Where; ET0 represents the evapotranspiration of plant (mm/day), Rn is the net radiation on vegetation cover (megajoul/m2 /day), T is the average air temperature (C), U2 introduces the wind speed at 2 m above the ground level (m/s), ea − ed shows the pressure shield shortage at 2 m (kPa), Δ is determined by the slope of the vapor pressure curve (kPa/C), γ is the psychometric coefficient (kPa/C) and G is the heat flux into the soil (Mj/m2 /day). The SPEI is presented by Vicente-serrano et al. (2011) and it has been used in numerous studies; also, calculation method involves the equilibrium of the climate and considers the role of the temperature in the drought evaluation. Previously, the palmer drought severity index (PDSI) uses the readily available temperature and precipitation data to estimate relative dryness (Palmer 1965). It is a standardized index that spans − 10 (dry)–+10 (wet) (Vicente-serrano et al. 2010). It has been reasonably successful at quantifying long-term drought. As it uses temperature data and a physical water balance model, it can capture the basic efect of global warming on drought through changes in potential evapotranspiration. Some studies have compared diferent methods for calculating PET (Shefeld et al. 2012) and it has been shown that the Penman–Monteith equation approximates net evapotranspiration (ET), requiring as input daily mean temperature, wind speed, relative humidity and solar radiation (Allen et al 1998). Therefore, in this study the calculation of SPEI is based on Penman–Monteith equation with Hargreaves-Samani correction that described in FAO56 (Allen et al. 1998). The PM method chosen by the World Meteorological Organization (WMO) that has been proposed as the standard method for PET estimation, and its accuracy has been proved by the need of less data. The monthly values of the reference plant evapotranspiration are calculated based on the climatological data and using Eq. (3). In the next step, the diference between the precipitation (P) and evapotranspiration (ET0) for the ith month can be calculated:

In order to calculate SPEI in diferent timescales, it is necessary to form series at different time steps (Paulo et al. 2003):

where wi,l is the value of Di or the water requirement in the frst month of the year (mm). The SPEI is used at various times for monitoring the agricultural and hydrological droughts (komuscu, 1999). The SPEI calculating, like the SPI method, requires estimating cumulative probability values of Di by fitting a probability function. Now, Di values from the lower bound lead to the negative values and two-parameter probability functions cannot be selected. In this research, with goodness-of-ft test, extreme value distribution function was selected for Di series and the drought analysis was performed on time steps of 3, 6, 9, 12, 24 and 48 months. The crossover probability values of Di series are converted to the standardize function with mean zero and standard deviation of 1, which will be equal to the values of SPEI (Vicente-Serrano 2006): In which P is the probability of exceeding the D and P = 1-F (x) values. If the value of P is greater than 0.5, then the value of P is replaced with 1-P and SPEI sign is changed.

Table 4. The relationship between drought severity and duration
Source: Pathak and Dodamani 2019, p 10

The RDI for the “i” year of month “k” will be calculated using:

3.6 Ethical Considerations

Some of the ethical considerations to make in the drafting of this paper, is the conflict of interest that may arise due to the subject of discussion. I declare that I absolutely have no conflict of interest in this subject matter. The paper is open to critique and use for scholarly and educational purposes and to offer insight into the possible measures that could be taken to rectify any and all damages to water resources in the UK. The images, tables and information adopted has been borrowed from other sources that have been clearly referenced. Therefore, there is no attribution of credit to the adopted works rather than to the original persons from whence adopted.

3.7 Assumptions, Delimitations, and Limitations

3.7.1 Assumptions

The major assumption perpetrated by this paper is that our study area – the Elmley Marshes is a candid reflection of the current state of things in the UK. Such a notion may not be entirely accurate but the significance of the results and the study may be reflected in some similar areas within the country due to nearness in geographical proximity. The idea that our methods of analysis and data collection over a period of time would remain accurate, is another major assumption that we made. Due to the limitations of human involvement in data collection and analysis, we assumed that our activities had little or no margin of error.

3.7.2 Delimitations

Our study was exhaustively conducted through online sources and was heavily reliant on previous journals and research done by experts in climate change related subjects. As such, we were able to cover majority of the bases in terms of a wide range of information to read from. With the varying case studies and deployment of various methods of analysis for different geographical areas, we learnt the best methods to use for different scenarios. Consequently, our research conjured up the realization of vast knowledge and understanding in research and development for scholarly purpose.

3.7.3 Limitations

Due to the fact that our research was conducted through journals, papers and articles written by other scholars, there is a limitation to the amount of information we collected. Had we conducted on-ground research for this subject matter for present results, we might have realized different results and therefore different appropriations for the future. There was also the limitation of the use of human-dependent methods of analysis. Since we were reliant on calculations for data analysis there was an allowable margin of error that we input. Even though it might be insignificant, the results spanning from the same may not have been truly accurate. Therefore, the deductions from this analysis may not be accurate as well.


Figure shows simulated groundwater depths from each of the climate change runs grouped according to the modified potential evapotranspiration data used (PETtemp and PETtrws). It also shows results from the calibrated model for the same MIKE SHE grids square and observations from a piezometer installed at this location. The simplicity of the hydrogeological conditions within the model and low gradients means that groundwater depths for each simulation are representative of those in the low-lying part of the model area. Under all the PETtemp scenarios the water table still rises to intercept the surface in the first two winters of the simulation period despite falling further at the end of the preceding summers. However, this rise is delayed, especially in 1998/1999. Table 3 summarises the number of days in each complete hydrological year when the simulated water table was above threshold depths close to the ground surface. In 1997/1998 the calibrated water table was within 0.01 m of the surface for nearly 70 days and within 0.10 m for nearly 120 days. However, under all four PETtemp scenarios the water table falls in mid-February and, although rising in March, does not intercept the ground again until mid-April. The number of days when the water table is close to the surface therefore declines. Although there is a general reduction in the number of high groundwater days in 1997/1998 from Ltemp to Htemp, the results for MHtemp are an exception. Increased winter precipitation results in small increases in net precipitation for Htemp compared to MHtemp despite higher evapotranspiration rates. This prolongs the period of high-water table for the former scenario compared to MHtemp.

In the winter of the following year, PETtemp water tables have to rise further still before intercepting the surface which they do between 40 (Ltemp) and 56 (Htemp) days after the calibrated results. Under the Htemp scenario the spring water table falls 37 days earlier compared to the other scenarios. The duration of periods when water tables are close to the surface declines from the calibrated results through each of the progressively higher emissions scenarios. Declines associated with Htemp are particularly large. Water tables during the following spring and summer are again lower for the PETtemp scenarios whilst the gains in groundwater elevation in autumn/winter of 1999/2000 are very subdued compared to calibrated results. Changes in groundwater depth are more pronounced for the PETtrws scenarios (Fig. 4b). Differences between results of the four climate change scenarios are particularly large in winter. Higher spring and summer evapotranspiration causes even lower groundwater levels at the start of autumn and, despite increases in winter precipitation, the water table fails to recover to the elevations evident in calibrated results. The water table rarely approaches the ground surface. Only under the Ltrws scenario does it reach within 0.10 m of the surface and then for only very short periods (Table 3). Under the Htrws scenario, the water table never comes within 0.20 m of the surface whilst it only does so under the MLtrws and MHtrws scenario in 1997/1998. Only very small gains in groundwater level occur in the autumn/ winter of 1999/2000 when calibrated results show rapid gains in water table elevation. Table 4 shows minimum, maximum and range of groundwater depths for the calibrated model and each climate change scenario for both complete hydrological years of the simulation period. Since the recession in water tables extends beyond the end of the usual hydrological year (September–August), this analysis defines the hydrological year as the period from the lowest water table before one autumn/winter rise to the lowest water table the following year.

Peak winter groundwater levels in all of the PETtemp scenarios in both years are the same since the water table intercepts the ground surface (Table 4 provides the depth of surface water when this occurs but since flooding results from groundwater intercepting the surface, ponding of precipitation and inundation from ditches the minimum groundwater depth is assumed to be 0.0 m). Consequently, lower summer water tables lead to small increases in the seasonal range of groundwater depths. In contrast, the seasonal range of groundwater depths declines from the Low through to the High emissions scenarios which use PETtrws. Groundwater changes are illustrated further by depth–duration curves for each scenario and the calibrated model derived using results from the complete simulation period (Fig. 5). Drier conditions for the progressively higher emissions scenarios are demonstrated as is the more extreme drying trend for the PETtrws scenarios. Increases in the range of groundwater depths for the PETtemp scenarios are evident (Fig. 5a) as is the virtual elimination of saturated conditions at the ground surface and the reduction in range of groundwater depths for the PETtrws scenarios (Fig. 5b).

Fig. Depth–duration b graphs for simulated groundwater depths within the Elmley Marshes for the calibrated model and UKCIP02 climate change scenarios for 2050: a. Using PETtemp, b. using PETtrws
Source: Thompson, Gavin, Refsgaard, Sørenson and Gowing 2009, p. 503-510

Ditch water levels Figure 6 shows simulated ditch water levels for the eight climate change scenarios grouped according to the potential evapotranspiration data used (PETtemp and PETtrws). Results from the calibrated model for the same location as well as a stage board installed in this ditch are also shown. For each scenario the results are representative of those throughout the MIKE 11 model in which water levels are approximately uniform due to the low gradients and interconnected nature of the ditch network. Figure 6a shows that for the four PETtemp scenarios a general lowering of ditch water levels occurs throughout most of the simulation period. Spring and summer drawdowns are larger for each of the progressively higher emissions scenarios. Despite increased precipitation between November and March, ditch water levels for all PETtemp scenarios fail to reach those experienced under calibrated conditions in the winter of 1997/1998 and the initial rapid rise is delayed by around 10 days. Whereas the calibrated ditch water level exceeds the elevation of the MIKE 11 weirs (1.75 m OD) on 40 days in 1997/1998, this does not occur once under any of the PETtemp scenarios. Although ditch water levels in 1998/1999 exceed the weir elevation under all the PETtemp scenarios the rise is progressively delayed (first reaching 1.75 m OD 17 and 40 days after the calibrated results for Ltemp and Htemp, respectively). Some short-lived peaks resulting from individual rain events which increase in magnitude under climate change exceed calibrated ditch water levels. Water levels begin to recede earlier in the spring of 1999 so that the number of days when the elevation of the weirs is exceeded declines. Under calibrated conditions ditch water would overtop the weirs on 78 days in 1998/1999. The corresponding figures for Ltemp and Htemp are 65 and 15, respectively. Towards the end of the simulation period, the rapid rise in ditch water levels shown in the calibrated results is delayed and reduced in magnitude under the progressively higher emissions scenarios and is absent in the Htemp results. Differences between calibrated ditch water levels and those associated with climate change scenarios are larger when using PETtrws and the overall drying trend is enhanced considerably (Fig. 6b). The magnitude of annual drawdown increases systematically from Ltrws through to Htrws and the duration of the drawdown increases from year to year so that the autumn/winter gains in water level are delayed. These increases in water level are of similar magnitude to those of the calibrated results in 1997/1998 whilst in the following year they exceed the rise in calibrated ditch water levels. However, the lower initial levels ensure that in both years ditch water fails to reach the elevation of the MIKE 11 weirs. Peak winter ditch water levels are very similar for each climate change scenario and each hydrological year. A rise in ditch water levels in the winter of 1999/2000 is absent from the results of all four PETtrws scenarios. The ditch water level–duration graphs of Fig. 7 summarise the changes resulting from the climate change scenarios. The overall lowering of water levels from the calibrated through the Low to the High emissions scenario for both the PETtemp and PETtrws simulations is demonstrated as are the larger reductions in ditch water levels associated with the PETtrws scenarios. In all cases the range in ditch water levels increases compared to calibrated results. For the PETtemp scenarios there is little difference in the range of ditch water levels between emission scenarios with the exception of Htemp which has a wider range due to the absence of any recovery in the winter of 1999/ 2000. In contrast, the overall ditch water level range increases progressively from Ltrws to Htrws.

Surface inundation

Inundation of the marsh surface results from a combination of high-water tables which intercept the ground surface especially within low-lying rills, movement of water from ditches often along these rills, and ponding of precipitation and local runoff (Thompson et al. 2004). The climate change scenarios reduce the incidence of conditions conducive to surface flooding by lowering groundwater and ditch water levels and by reducing the duration of periods of high-water level when they do occur. Figures 8 and 9 show the extent of open water for a range of maximum depths for the PETtemp and PETtrws scenarios, respectively. The corresponding data for the calibrated results are also shown. The extent and duration of shallow (maximum depths 0.1 and 0.2 m) flooding under the PETtemp scenarios (Fig. 8a, b) reflect changes in both groundwater and ditch water levels. The area of shallow flooding is reduced whist the progressively longer delay in the expansion of flooding in 1998/1999 results from delayed gains in groundwater and ditch water levels. Similarly, the earlier recession in water levels cause shallow inundation to recede earlier. Depleted groundwater levels between mid-February and mid-April 1998 reduce inundation at this time with the middle peak in flooding of 0.2 m maximum depth shown in the calibrated results being noticeably absent (Fig. 8b). Some shallow inundation is retained through most of the spring and summer although its extent is reduced. Figure 8c shows that in 1997/1998 the one distinctive peak in flooding of 0.3 m maximum depth shown in the calibrated results is absent from all the PETtemp scenarios. Thompson et al. (2004) suggested that a threshold ditch water level of 1.75 m OD, coupled with a high-water table, is associated with the expansion of flooding of this depth. This threshold was not exceeded in 1997/1998 for any of the PETtemp scenarios. In the following year the threshold was exceeded under all the PETtemp scenarios although some time after the calibrated results. Some small winter peaks exceeded those of the calibrated results. Figure 8c shows short-lived periods for three of the PETtemp scenarios (Ltemp, MLtemp and MHtemp) when the extent of flooding of 0.3 m maximum depth exceeds that shown for the calibrated results although the maximum extent of flooding of this depth is still associated with the calibrated model. On three occasions in 1998/1999 the extent of flooding of 0.4 m maximum depth under calibrated conditions is exceeded by some climate change scenarios (Ltemp and/or MLtemp) (Fig. 8d). However, these increases are small and the predominant trend is for progressively smaller areas of deeper flooding from Ltemp to the Htemp. In 1997/1998 the extent of areas flooded to a maximum depth of 0.4 m is generally consistent between PETtemp scenarios and is lower than under calibrated conditions, when these areas were already small and restricted to locations immediately adjacent to ditches in the far south of the marshes. The much larger changes in groundwater and ditch water levels for the PETtrws scenarios lead to dramatic reductions in flood extent. Figure 9a shows that in 1997/1998 there are three peaks in shallow (0.1 m) flooding of approximately the same magnitude but much reduced compared to the calibrated results. Not all of these peaks are present in results of the four scenarios. Both the Ltrws and MLtrws scenarios yield the first, only Ltrws provides the second whilst all four scenarios display the third peak. Thompson et al. (2004) identified a threshold ditch water level of 1.57 m OD required to initiate shallow inundation. High groundwater levels also promoted inundation although a distinct threshold water table depth was less discernible. Figure 6b shows three distinct peaks in ditch water levels for the PETtrws scenarios during the 1997/1998 period of elevated water levels. Those which exceed the 1.57 m OD threshold for more than a day or two correspond with the peaks in shallow flooding (Fig. 9a). In 1998/1999 the extent of shallow flooding is small and consistent between the PETtrws scenarios. Its maximum extent is less than 10% of the corresponding area for calibrated results. Ditch water levels for most of the PETtrws scenarios failed to reach the 1.57 m OD threshold in 1998/1999 and for the one scenario where they did (Ltrws) levels only exceeded this threshold by a very narrow margin and for short periods. Figure 9a shows that in the spring and summer flooding of 0.1 m depth disappears completely. All of the PETtrws scenarios have eliminated significant areas flooded to a maximum depth of 0.2 and 0.3 m (Fig. 9b, c). Ditch water levels fail to reach the 1.65 and 1.75 m OD thresholds identified by Thompson et al. (2004) which are required for flooding of these depths. Similarly, the extent of deeper (0.4 m) inundation is further restricted under the PETtrws scenarios (Fig. 9d). The peak extent of inundation of this depth varies little between scenarios, implying that the same areas are flooded, although the duration of its presence tends to be longer for the lower emissions scenarios.


The use of drought indices allowed studying and analysis of the characteristics and propagation behaviours of droughts using SPI, SPEI and RDI for extracting the time series of dry and wet periods during past periods. From the computation and analysis of these indices at meteorological and hydrometric stations in study areas, we observed that droughts were depicted by these indices based on the meteorological and hydrological data. The SPEI is more appropriate than the SPI for applications examining of drought variability in study areas because it uses precipitation and evapotranspiration data. The results showed that precipitation plays a significant role in explaining the time variations of the droughts and temperature increasing also has a significant effect on water shortages, especially in long-term periods in the region. The Pearson correlation matrix and crossover correlation show that SPI drought index is better than SPEI drought index in the Elmley Marshes since it is a wet zone and with less evapotranspiration. However, the SSI hydrological drought index responds faster to SPEI drought index as compared to SPI. In the arid and semiarid areas, because of much evapotranspiration, the SPEI is better than the SPI and it was found that RDI has responded to both SPI and SPEI in the watershed basin. Extracting the nonlinear model between the meteorological and hydrological drought shows that there is a relationship between the hydrological drought response to the meteorological drought. In general, the application of several indices indicating different components of the hydrological cycle integrates many factors that affect and trigger droughts, and thus can help in providing a wider realization of the characteristics of droughts on various water sections.

Suggestions for Future Work

For any future work to be undertaken as appertaining to UK water resources due to climate change, I would suggest that greater focus be taken on the highland areas of the UK. It will enable us to have an overview of both extremities in the country and therefore further assist in deriving conceivable conclusions that may be more accurate than what we currently have.

Reference List

Adas, A., 1997. Traffic models in broadband networks. IEEE communications Magazine, 35(7), pp.82-89.

Alcamo, J., Flörke, M. and Märker, M., 2007. Future long-term changes in global water resources driven by socio-economic and climatic changes. Hydrological Sciences Journal, 52(2), pp.247-275.

Al-Khudhairy, D.H.A., Thompson, J.R., Gavin, H. and Hamm, N.A.S., 1999. Hydrological modelling of a drained grazing marsh under agricultural land use and the simulation of restoration management scenarios. Hydrological Sciences Journal, 44(6), pp.943-971.

Gavin, H., 2001. Wetland hydrology of the Elmley Marshes (Doctoral dissertation, University of London).

Wilson, A.M., Ausden, M. and Milsom, T.P., 2004. Changes in breeding wader populations on lowland wet grasslands in England and Wales: causes and potential solutions. Ibis, 146, pp.32-40.

Ausden, M. and Hirons, G.J.M., 2002. Grassland nature reserves for breeding wading birds in England and the implications for the ESA agri-environment scheme. Biological Conservation, 106(2), pp.279-291.

Barthel, R., Reichenau, T.G., Krimly, T., Dabbert, S., Schneider, K. and Mauser, W., 2012. Integrated modeling of global change impacts on agriculture and groundwater resources. Water Resources Management, 26(7), pp.1929-1951.

Davies, E.G. and Simonovic, S.P., 2011. Global water resources modeling with an integrated model of the social–economic–environmental system. Advances in water resources, 34(6), pp.684-700.

Flato, G., Marotzke, J., Abiodun, B., Braconnot, P., Chou, S.C., Collins, W., Cox, P., Driouech, F., Emori, S., Eyring, V. and Forest, C., 2014. Evaluation of climate models. In Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 741-866). Cambridge University Press.

Fordham, S.J. and Green, R.D., 1980. Soils of Kent [England]. Soil Survey Bulletin (UK). no. 9.

Gain, A.K., Giupponi, C. and Renaud, F.G., 2012. Climate change adaptation and vulnerability assessment of water resources systems in developing countries: a generalized framework and a feasibility study in Bangladesh. Water, 4(2), pp.345-366.

Green, R.E. and Robins, M., 1993. The decline of the ornithological importance of the Somerset Levels and Moors, England and changes in the management of water levels. Biological Conservation, 66(2), pp.95-106.

Hattermann, F.F., Post, J., Conradt, T. and Wechsung, F., 2008. Assessment of water availability in a Central-European River Basin (Elbe) under climate change. 气候变化研究进展, 4(000), p.42.

Havnø, K., Madsen, M.N. and Dørge, J., 1995. MIKE 11-a generalized river modelling package. Computer models of watershed hydrology., pp.733-782.

Manifold, L., del Strother, P., Gold, D.P., Burgess, P. and Hollis, C., 2021. Unravelling Evidence for Global Climate Change in Mississippian Carbonate Strata from the Derbyshire and North Wales Platforms, UK. Journal of the Geological Society.

Jefferson, R.G. and Grice, P.V., 1998. The conservation of lowland wet grassland in England. European Wet Grasslands: Biodiversity, Management and Restoration, pp.31-48.

Joyce, C.B. and Wade, P.M., 1998. European wet grasslands: biodiversity, management and restoration. John Wiley and Sons Ltd.

Kelman, I. and Gaillard, J.C., 2010. Embedding climate change adaptation within disaster risk reduction. In Climate change adaptation and disaster risk reduction: Issues and challenges. Emerald Group Publishing Limited.

Mourato, S., Moreira, M. and Corte-Real, J., 2015. Water resources impact assessment under climate change scenarios in Mediterranean watersheds. Water Resources Management, 29(7), pp.2377-2391.

Murray, S.J., Foster, P.N. and Prentice, I.C., 2012. Future global water resources with respect to climate change and water withdrawals as estimated by a dynamic global vegetation model. Journal of Hydrology, 448, pp.14-29.

Muzik, I., 2001. Sensitivity of hydrologic systems to climate change. Canadian Water Resources Journal, 26(2), pp.233-252.

O’Hara, J.K., 2007. Water resources planning under climate change and variability (Doctoral dissertation, UC San Diego).

Olmstead, S.M., 2014. Climate change adaptation and water resource management: A review of the literature. Energy Economics, 46, pp.500-509.

Olsen, A.S., Zhou, Q., Linde, J.J. and Arnbjerg-Nielsen, K., 2015. Comparing methods of calculating expected annual damage in urban pluvial flood risk assessments. Water, 7(1), pp.255-270.

Pathak, A.A. and Dodamani, B.M., 2019. Comparison of meteorological drought indices for different climatic regions of an Indian river basin. Asia-Pacific Journal of Atmospheric Sciences, pp.1-14.

Pingale, S.M., Jat, M.K. and Khare, D., 2014. Integrated urban water management modelling under climate change scenarios. Resources, Conservation and Recycling, 83, pp.176-189.

Praskievicz, S. and Chang, H., 2009. A review of hydrological modelling of basin-scale climate change and urban development impacts. Progress in Physical Geography, 33(5), pp.650-671.

Randall, D.A., Wood, R.A., Bony, S., Colman, R., Fichefet, T., Fyfe, J., Kattsov, V., Pitman, A., Shukla, J., Srinivasan, J. and Stouffer, R.J., 2007. Climate models and their evaluation. In Climate change 2007: The physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the IPCC (FAR) (pp. 589-662). Cambridge University Press.

Rodwell, J.S. and Joint nature conservation committee (GB), 2006. National vegetation classification: Users’ handbook. Peterborough: Joint nature conservation committee.

Salimi, H., Asadi, E. and Darbandi, S., 2021. Meteorological and hydrological drought monitoring using several drought indices. Applied Water Science, 11(2), pp.1-10.

Thompson, J.R., Gavin, H., Refsgaard, A., Sørenson, H.R. and Gowing, D.J., 2009. Modelling the hydrological impacts of climate change on UK lowland wet grassland. Wetlands Ecology and Management, 17(5), pp.503-523.

Vishwakarma, A., Choudhary, M.K. and Chauhan, M.S., 2020. Applicability of SPI and RDI for forthcoming drought events: a non-parametric trend and one way ANOVA approach. Journal of water and climate change, 11(S1), pp.18-28.

Watts, G., Battarbee, R.W., Bloomfield, J.P., Crossman, J., Daccache, A., Durance, I., Elliott, J.A., Garner, G., Hannaford, J., Hannah, D.M. and Hess, T., 2015. Climate change and water in the UK–past changes and future prospects. Progress in Physical Geography, 39(1), pp.6-28.

Stuart, M.E., Gooddy, D.C., Bloomfield, J.P. and Williams, A.T., 2011. A review of the impact of climate change on future nitrate concentrations in groundwater of the UK. Science of the Total Environment, 409(15), pp.2859-2873.

Stockle, C.O., Williams, J.R., Rosenberg, N.J. and Jones, C.A., 1992. A method for estimating the direct and climatic effects of rising atmospheric carbon dioxide on growth and yield of crops: Part I—Modification of the EPIC model for climate change analysis. Agricultural systems, 38(3), pp.225-238.

Acreman, M.C., Blake, J.R., Booker, D.J., Harding, R.J., Reynard, N., Mountford, J.O. and Stratford, C.J., 2009. A simple framework for evaluating regional wetland ecohydrological response to climate change with case studies from Great Britain. Ecohydrology: Ecosystems, Land and Water Process Interactions, Ecohydrogeomorphology, 2(1), pp.1-17.

Yousefpour, R., Temperli, C., Jacobsen, J.B., Thorsen, B.J., Meilby, H., Lexer, M.J., Lindner, M., Bugmann, H., Borges, J.G., Palma, J.H. and Ray, D., 2017. A framework for modeling adaptive forest management and decision making under climate change. Ecology and Society, 22(4).

Zhou, Q., Mikkelsen, P.S., Halsnæs, K. and Arnbjerg-Nielsen, K., 2012. Framework for economic pluvial flood risk assessment considering climate change effects and adaptation benefits. Journal of Hydrology, 414, pp.539-549.

Zhou, Y. and Guo, S., 2013. Incorporating ecological requirement into multipurpose reservoir operating rule curves for adaptation to climate change. Journal of Hydrology, 498, pp.153-164.

Leave a Reply

Your email address will not be published.