The aim of thisresearch is to develop a conceptual MJ-GIS model that can be uniformly applied in support of regionalised emergency disaster management (EDM). It seeks to achieve this by focusing on the potential synergies of Local and Regional Authorities and their collective ability to acquire, maintain and disseminate spatial data specific to the needs of regionalised EDM.
The literature review identified that emergency management activities can generally be grouped into four distinct, yet often overlapping phases each related by time and function (ESRI 1999). It also identified four distinct GIS models commonly used by Local and Regional Authorities to support EDM and briefly outlines the major benefits and challenges of each. Finally it reviews the more common problems related to organisational and inter-agency data sharing.
A survey was then conducted to measure the occurrence of broadly comparable experiences with those identified in the literature review. While some of the results proved to be varied and inconclusive, others were found to support the literature review and verified that the most significant challenges to sharing data between independent agencies are more closely related to institutional, organisational and behavioural issues, rather than technical ones.
It is believed that the conceptual MJ-GIS model presented in this dissertation will address many of these widely experienced challenges and that it is theoretically possible to implement it successfully anywhere in the world. However by also acknowledging that technical issues are much more easily solved than human ones, its success will ultimately depend upon the willingness of its participants to collectively work together in overcoming them.
Though many organisations have introduced organisation-wide Geographical Information Systems (GIS), very few are obtaining the full benefits these systems offer. The British Government have a key role to play in leading the use of GIS, both as a main data supplier and a large-scale user. If leading large-scale Government Departments are struggling to extend the use of GIS beyond a mapping functionality, there is little hope for others.
Barriers restricting the use of GIS are often associated with behavioural, cultural and organisational factors. This qualitative study supports this thinking further using an in-depth investigation of a leading Central Government Department using geographic information. Generalisations from a case study analysis with two other organisations are turned into best practice principles. It is hoped these principles will help other organisations in similar circumstances.
Organisational change is found to be essential for improving the use of GIS and spatial analysis in organisations. It is recommended more emphasis needs to be made to developing and delivering an internal communications plan to raise awareness of GIS more widely across the organisation. A bottom-up approach to link the use of GIS to business objectives, rather than a traditional top-down strategic approach is also recommended This should be supported by obtaining a view of business areas and staff within the organisation that can be expected to use or develop GIS and spatial skills.
(no abstract within his MSc)
Southern California is a seismically active region with a large population. The Study Area is the Fontana Seismic Trend, an area of intense earthquake activity in western San Bernardino County . There has been not been much significant work done to quantify the potential seismic hazard posed by it. This dissertation will show that there is potential for a moderate seismic event on the Fontana Seismic Trend that will likely cause quite a bit of damage and disrupt lives. The major faults nearby have been studied extensively, but this is not the case for the Fontana Seismic Trend. When plotted on a map, the earthquake epicenters of the Fontana Seismic Trend display a significant lnear pattern both two dimensionally and three dimensionally. In this dssertation, the location and direction of movement of the underlying fault were identified and configured to fit the earthquake data. Using the plotted fault location, direction of movement and length, the maximum potential event that might occur on this fault was also quantified. HAZUS-MH (a FEMA software extension to ArcGIS) was used to perform a seismic hazard analysis to find out what type of damage and casualties can be expected if a moderate event occurs on this fault. This dissertation discusses the methods used to determine the fault parameters and details the results from the seismic hazard analysis.
After decades of decline urban greenspaces are back on the political agenda as they are seen as improving health, social inclusion, community cohesion and safeguarding local wildlife and nature. Local authorities, following Government policy, are carrying out audits of greenspace provision in terms of quantity, quality and accessibility and, due to the spatial nature of these audits, the use of GIS is highly recommended. The Government believe standards should be set locally, this has resulted in varied and inconsistent approaches which hinder the additional Government priorities of data sharing and benchmarking. To determine if there is sufficient quantity of greenspaces and if those of a higher quality are distributed equitably then accessibility is the key issue. The two main methods for measuring accessibility use buffer and network analysis and these were investigated, based on variable walking distances, utilizing data supplied by the local authorities of Trafford and Manchester . Local authorities have favoured the buffer method because network analysis is considered complicated, however, this study has shown the buffer method considerably overestimates accessibility compared to the network method. Local authorities have also tended to ignore cross boundary accessibility which in the case of Manchester ’s boundary with Trafford has been shown to be sizable. Finally, t he analysis of accessibility to high quality greenspaces using socio-economic census data highlights the need to carryout comprehensive quality audits of all greenspaces to ensure equity is present and funding is being distributed evenly. If the Government is going to achieve its aim of an urban renaissance it needs to reconsider its urban greenspace policies and seriously consider a comprehensive national standard to cover all aspects of quantity, quality and accessibility.
Qualitative research is a rigorous method that allows the capture of rich and complex meanings from its subjects. In the case of this dissertation, qualitative analysis made sense because GPS has never been fully implemented at BC Hydro and the foundation for evaluating such technology is grounded in the experiences of those who design and install the workings of an electrical distribution network. The qualitative analysis and the resulting grounded theory used in this research was the major driving force in determining exactly what equipment within an electrical distribution network would benefit from a GPS survey.
By providing underground and overhead electrical equipment their own UTM coordinate within the GIS, locating this equipment at a later date for maintenance or design reasons is as simple as using a process called “reverse GPS”. The user can navigate back to the original UTM coordinate and be within centimeters of the original location thereby reducing the labour associated with locating electrical equipment in any weather condition or time of day.
GPS technology can only be successfully implemented if the cost benefits outweigh existing design methods. With an original investment of $19,600 for a good quality commercial grade GPS unit, the pay back period would be just under one year with surveying of existing roads getting the greatest cost benefit return.
In many villages, groundwater is used as a primary source for drinking water. To access the aquifers containing the water, wells are drilled from the surface to the future drinking water source. These private water supplies generally serve fewer than a few houses. As a result, a village will have many wells drilled in close proximity to each other. This close arrangement coupled with the sheer number of wells provides avenues for contaminants to pollute the water supply.
For larger single municipal supply wells servicing towns, the landbase around the well has limitations placed on what chemicals, pathogens and landuse can occur. The landbase is delineated into wellhead protection area zones with more stringent limitations enforced the closer one is to the well. The delineation process is completed through a number of different modelling techniques; each method with its own unique properties. Methodologies for applying the same principles to village wellfields do not exist.
This dissertation reviews 4 delineation methodologies: arbitrary fixed radius, calculated fixed radius, uniform flow method and finite element flow numerical modelling, for creating wellhead protection areas for individual wells within a village wellfield as well as investigates a single central well modelling methodology.
The results showed that the zones around the wells are in the order of 100m or less for a fixed time of 250 days. Overlap between adjacent zones exists for the arbitrary fixed radius method, less so for the calculated fixed radius method. The remaining two methods create very localised protections areas.
Due to uncertainty in the parameters and the use of the cautious approach in the use of modelling results, the preferred method for village wellfield wellhead protection area delineation was found to be a circular zone inclusive of all village wells. Additional zones were then radiated from the edge based on groundwater flow times.
A review of GIS literature – both past and present – shows numerous voices raised in criticism of Geographic Information Systems as a tool for anything more than the most rudimentary of spatial analysis. Some advantages – such as superior visualisation abilities – are grudgingly conceded, but this block of negative opinion argues that in the case of complex spatial analyses, the costs of using GIS far outweigh any potential gains.
Spatial ETL (Extract, Transform and Load) is a field of data warehousing whose applications are designed to read any number of disparate data sources, manage their integration and processing, and to output results to any desired destination. This type of functionality would appear to hold clear benefits in the domain of spatial data analysis, where one might cynically observe that data formats and standards appear to proliferate at a faster rate than interoperability considerations can keep up to. At the very least, one might expect Spatial ETL to be the ideal framework upon which to hang more advanced analysis tools.
However, to date, Spatial ETL is a largely ignored technology. There is no formal definition of the term, and no set of rules to classify what sort of functions a Spatial ETL application ought to include. Neither is there any serious academic debate about the subject, perhaps indicative of a blinkered view that GIS is the be-all and end-all of spatially related technologies.
By creating a comprehensive definition of Spatial ETL, including a list of rules characterizing Spatial ETL applications, this research shows where such tools can fit into the process of geographic analysis, and what value can be achieved by their use. Furthermore, a comparison of Spatial ETL and GIS applications highlights the benefits to be gained by using such tools in harness with one another.
In 2000, the UK Government launched their 10 year strategy aimed at reducing the number of casualties on Britain ’s roads. In 2004, the Government removed the ‘National Standards of Fire Cover’, against which all UK Fire and Rescue Services (FRS) had been measured since 1958, replacing them with a requirement for Fire Authorities to develop local standards, aimed at reducing the risk to life in their area. That same year the new Fire and Rescue Service Act came into force which required all FRS to plan for and respond to road traffic collisions (RTC). Consequently, FRS are becoming increasingly involved in supporting their local Road Safety Partnership’s efforts to reduce RTC’s.
Despite significant success with the Government’s strategy, m ore than 30,000 people continue to be killed or injured on Britain ’s roads each year. Recognising the continuing need for an effective response to each and every one of these emergencies, many FRS are now developing response standards specifically targeting these incidents.
Following a review of existing RTC risk assessment methods, used by FRS and other organisations, various GIS statistical techniques are used to explore the potential for the primary source of RTC data in the country (STATS19) to be used as the basis for Shropshire Fire and Rescue Service’s new response standards. A simple, network based, risk assessment method is proposed which has the potential to improve on this case study FRS’s current ability to plan its RTC response and prevention activities. Feedback obtained from professionals in this FRS and the other road safety organisations also indicates that, if implemented, the method will assist in communications between the FRS and its partners. With appropriate validation, the method could support other FRS in a similar way.
Performance Assessment of GIS implementations is essential for the success of those installations. This is more important in government organizations due to social, economic and political reasons. Many quantitative methods in vogue like NPV, ROI, measure performance against economic considerations. This is difficult in government organizations due to the social benefits that they offer which are difficult to quantify.
The major aim of this study was the identification of performance indicators which can be used in assessing the performance of an operational non-commercial GIS organization.
The study defines an approach to identify the core issues governing the use of GIS in an organization. The method was tested on a case study which was a municipality. Core problems that exist in the case study municipality were identified and along with a study of organizational objectives and stakeholder expectations, key performance objectives were arrived at. These were used in identifying key performance Indicators. The study found that the core issues were common across many sections and were related to data, technology and services.
Municipalities are under pressure to improve the services they provide to companies and citizens. An efficient information exchange is a requirement for supporting local processes such as e.g. the spatial planning process.
At the European level, INSPIRE is being developed but the question is whether the municipalities are willing and able to implement a Spatial Information Infrastructure?
Theory shows that organisational rather than technological aspects will determine whether the municipalities will be able to implement a Spatial Information Infrastructure (SII). Based on the key factors for SII implementation, this research has developed a model to determine SII maturity of the studied municipalities from an organisational perspective. Besides, different models were applied to measure whether the municipalities are willing to implement SIIs as well as to measure the current implementation status in the municipalities.
Based on these three studies, we have determined whether the municipalities are willing and able to implement a SII.
The conclusion is that municipalities in the Dutch province Limburg (as a whole) are not willing and/or able to implement a Spatial Information Infrastructure. Research in the adjacent German Kreis Heinsberg yields similar results.
Indeed, the principal obstacles when implementing a SII have turned out to be of an organisational rather than of a technological nature. Finance and culture were important aspects.
At the same time, municipalities had insufficient knowledge of what a SII is and what it can mean for the municipalities.
Particularly the smaller municipalities have difficulties to implement a SII. Collaboration, possibly in a shared service centre, could be a solution to share the necessary knowledge and experience and gain efficiency. For ensuring the success of INSPIRE also at the local level, actions should be taken now in view of the detected bottlenecks!
GIS applications are commonplace in local government for the collection and recording of land ownership but their use as a tool to aid in the strategic management of UK local government assets is an area of current development. The role and function of local government is changing and the design and development of information systems to support a structured clear process for policy making, evaluation and simulation is becoming necessary to fulfil government expectations.
The results of this study show that GIS is currently used in asset management mainly for the recording and display of asset management datasets, attribute and location enquiries and plan production. Local government can develop a property strategy that aids service delivery by adopting an information management approach using corporate integrated database/GIS technology to enable the geographical analysis of property information and the development of performance indicators to inform decision-making and the asset management plan and to provide authority-wide access.
This research paper concludes with a model for using GIS within a local government asset management context, to aid land and property related decision-making through the examination of the process, data, methods and techniques required.
The theme of the study is GIS Interoperability and exchanging GIS data. The study examines GIS interoperability issue that affect or constrain the exchange of GIS data between utility companies in the study region. Using a questionnaire the study examines interoperability and data management related issues. By analysing the results and respondents comments the study establishes if there is willingness by utility companies in the region to exchange GIS data in the future or if they are prevented from participating because of interoperability issues. It looks at justifying future data sharing initiatives using Cost Benefit Analysis (CBA) by analyzing the cost and benefits associated with using GIS data instead of a conventional paper based approach. The literature review describes how the international GIS community are addressing the issues of GIS interoperability through the development of standards and GML. Through a case study analysis the experience of researches tasked with sourcing and managing GIS data for a flood management application highlights the difficulties associated with integrating GIS data from disparate sources. By importing sample GIS data interoperability issues are demonstrated while simultaneously justifying the need to re-establish a GIS data sharing initiative.
The latter half of 2006 saw the culmination of Central Government’s Improving Electronic Government initiative and with it the fruition of many Local Authorities’ e-government integration projects. The National Land and Property Gazetteer formed the basis for many of the e-government targets expressed in the Initiative. At the same time, the Ordnance Survey announced that a new address layer (Address Layer 2™) would be added to its MasterMap™ product. A conflict has therefore arisen between parties involved in address maintenance procedures, and disputes regarding intellectual property and freedom of data relating to this look set to be long-running. At this critical juncture in the development of a definitive address database in England and Wales , this study seeks to evaluate and define the factors which will determine the success or otherwise of attempts to provide this definitive national address dataset.
A survey addressed to Local Authority LLPG Custodians highlights the factors considered significant by them in the sharing of address data, and seeks to identify whether initiatives involving LLPGs really are, as hoped, the key to the effective sharing and integration of address data. An attempt is made to define a model of best practice for the updating of address change intelligence between relevant bodies.
The aim of the thesis is to propose a new Weighted Evidence Geographic
Information Systems (GIS) Approach called the Multivariable Hazard Quotient GIS Approach (MVHQ) to delineate and estimate the volume of contaminated soil. The approach is a possible solution to dealing with anthropogenic filled sites, in which …contamination is difficult to delineate and characterize by conventional methods.
The Weighed Evidence GIS Approach utilizes hazard quotient coefficients (HQ)to compare multiple contaminants and characterize the maximum hazard level on site. Using an adequate model of spatial continuity and by applying kriging interpolation procedures, the MVHQ GIS Approach reduces the number of variables in the evaluation to a few major contaminants or to a single maximum value, simplifying decision making for programs of remediation.
The Multivariable Hazard Quotient GIS Approach (MVHQ) has been tested on the example of a Case Study, which presents a process of evaluation of a contaminated site, starting with data validation and evaluation, structural analysis, and concluding with the remediation program and recommendations.
Public response to mineral planning applications in England is generally hostile and can cause significant problems for mineral operators at a late stage in their site acquisition and development cycle. Evidence suggests that the public reaction is borne as much out of lack of knowledge of the mineral planning system as any ‘NIMBY’ response.
Recent changes to the planning regime in England impose a greater requirement for mineral planning authorities (MPA) to consult with the public and to conduct Strategic Environmental Assessments (SEA) of their Mineral and Waste Development Frameworks (MWDF). GIS are used widely by the MPAs as Spatial Decision Support Systems (SDSS) in producing their development plans but, as a participatory tool, could be used as a means of encouraging greater public participation and increasing their understanding of the mineral planning process. Despite improvements, public participation in local government affairs remains at Arnstein’s level of ‘tokenism’ with the public having little real influence on the decision making processes.
A combination of commercial sensitivity, the nature of the planning system and of land and mineral ownership considerations means that the minerals industry is, and is likely to remain, reluctant to release more information into the public domain than is necessary to secure planning permission. Little use is made of GIS systems despite their ideal applicability to the estates and development functions of these companies.
Although this research confirms that P-GIS may present an opportunity to promote contributory participation, the intensive use of data by such systems and the general public’s current high level of disengagement from local democracy suggest that P-GIS will remain an expert tool for the foreseeable future.
A number of government agendas and the promise of additional funding is forcing many Local Planning Authorities (LPAs) to radical re-evaluate the use of websites for delivering services. Evidence indicates in an attempt to work smarter and more efficiently, LPAs are transforming the way services are delivered based around the needs of their customers with assistance from ICT and guidance material in the form of toolkits.
Pre-application advice is seen by many LPAs as an integral element of the planning process and is one area that is witnessing change for both officers and customers alike. For a small number of LPAs, the introduction of a web based planning permission enquiry system is already delivering a number of benefits, including significantly modernising and e-enabling the manner in which customers approach them for advice.
As with any decision-based system that depends wholly on data, the accuracy is extremely important. Incorrect decisions due to inaccuracies will lead to distrust and eventually to the possible demise of these genuinely innovative solutions. Clearly, as a result of findings provided as part of this research, a major area of concern for LPAs has been the accuracy and completeness of spatial data. Therefore, service providers must make a concerted effort to ensure spatial data is accurate at all times with attention focussed on and around the impact of positional accuracy improvement programme.
The research paper will conclude with a brief overview of several fundamental elements that findings suggest may impact on the success of these systems, including GIS data, system and data costs and ensuring effective take-up and continuing use of online services.
The identification of the fastest walking route between points in mountainous terrain is of interest to the military, mountain rescue teams, hill walkers, competitive orienteers and fell runners. The process is usually carried out using a topographical map. Factors such as the distance and the vertical ascent of different routes are taken into consideration in deciding on the fastest option. This process relies on the hill walking experience of the route planner to choose routes which avoid areas likely to provide slow walking conditions due to the land cover. The aim of this research is to investigate how remote sensing data and GIS can be used to improve the identification of such fastest routes. The study area used was a 25 square kilometre area to the north west of Kentmere in the English Lake District. The data sources used were a digital elevation model (DEM) to model the topography of the land, Landsat ETM images used to classify land cover which affects walking speed and digitised footpath data. The data was processed and analysed using a direction dependent function (VARCOST) in the Idrisi Andes GIS package. The final model developed can successfully identify fastest walking routes. The routes identified take ‘sensible’ and realistic paths which correspond with conventional hill walking experience in that the routes avoid steep and rocky ground if possible, in flatter areas the routes choose ground with land cover which is the easiest to traverse and footpaths are used when appropriate. The model can also provide an estimate of time required to walk the identified fastest routes.
UK Local Government is responsible for the delivery of many vital services such as education, housing and planning. It is constantly looking to improve service delivery. One way is through the use of technology. GIS (Geographic Information Systems) is one element of technology that is currently used, primarily through the provision of public facing mapping web pages.
The majority of GIS software used within Local Government is proprietary with ESRI and MapInfo dominating. Free/Open Source Software (FOSS) is rarely used. The aim of this dissertation is to investigate whether Free/Open Source web mapping software is a viable option for UK Local Government when re-evaluating their GIS service provision.
The question was investigated using three methodologies:
- A survey of UK Local Government GIS professionals.
- An analysis of the possible technical barriers to the adoption of Free/Open Source web mapping server software.
- The creation of a test mapping website using FOSS products which replicates the functionality commonly found on Local Government mapping web pages.
The analysis of the possible technical barriers and the creation of the test mapping website led to the conclusion that technical issues do not present a barrier to the adoption of Free/Open Source web mapping software.
The analysis of the survey results revealed that organisational issues present a significant barrier to the adoption of Free/Open Source web mapping software. The organisational issues are:
- Most local authorities outsource their IT solutions. However, there is a lack of external consultants who offer Free/Open Source web mapping solutions.
- GIS software within local authorities is dominated by proprietary products.
- The adoption of FOSS products entails greater risk within UK Local Government.
The catalyst for increased adoption of FOSS products by local authorities would be a more proactive approach by Central Government. This would include:
Clear Policy Initiatives
Clear and positive guidelines would be a stimulus for local authorities to investigate and adopt FOSS. This would create market forces for consultants to offer solutions based on FOSS. This could create a positive feedback situation with increased adoption stimulating the growth of a consultancy market and vice-versa.
Increased Funding
Increased funding of FOSS research and advisory bodies for local government would provide a source of external expertise. This would reduce risk.
This raises the question as to whether UK Central Government should pro actively promote FOSS more. The author believes the answer is yes as it will increase the range of software products available to UK Local Government. The subject is an interesting research project in its own right.
This study explores the feasibility of various techniques to detect, analyse and validate geological lineaments derived from remote sensed images. It attempts to utilize public domain data and readily available software to add or improve structural knowledge using lineaments. Hough Transform (HT) has been chosen as the technique for extracting local lineaments using aerial photographs and radar imagery. Results are analysed in vector line format using a custom built strip-based grouping application.
Human expertise, supplemented with prior knowledge of the local scale is needed to interpret and validate the results. Validated lineaments are then integrated with geological and geochemical data for input into a weights of evidence model for mineral potential mapping.
The automated lineament extraction results at the local scale performed much better at detecting geologically significant lineations, being robust against various terrains and poor photographic exposures. The weights of evidence model predicted several areas at the regional scale that may warrant further exploration, while at the local scale weights of evidence performance was less successful. Predicted results from both manual and automated lineament extraction results correlated to a high degree, suggesting that automated lineament interpretations is a valid method in lineament studies
This research project involved creating an automated mapping system to produce maps for use in the 2007 Census Test in England and Wales . The methods used in the 2001 UK Census were evaluated, and a review of relevant literature carried out to find similar systems. The reasons for automating the map production process were considered, requirements for the content and style of the maps were identified, and a system created using ESRI’s ® ArcGIS Desktop products and Visual Basic for Applications macros.
The success of this approach was evaluated using qualitative and quantitative techniques using a voluntary questionnaire provided to field staff, the information provided during debriefing sessions carried out after the 2007 Test, and speed tests for producing maps using different techniques. Results from the questionnaire, debriefing, and speed tests indicate that the mapping system was a success and can provide the basis for the 2011 Census mapping system. With this in mind, suggestions for future improvements have been provided.
The spatial distribution of population in cities varies widely throughout the day as people travel between their homes and places of work, education, retail and leisure. Knowledge of where people are located at different times may assist emergency responders in the event of the disaster. Authorities also need to prepare evacuation plans based on potential scenarios.
The development of models to simulate the diurnal variation of population distribution in urban areas has been constrained by data availability. Many existing models therefore make gross assumptions about daytime population distribution based on limited and spatially coarse census data. The aim of this dissertation is to research the theoretical development of a diurnal population model and to devise a prototype software application to calculate the distribution of urban population at different times. This study has practical relevance to emergency planning as it enables simulations to be generated that may be used in contingency planning.
The concept of temporal profiles is proposed where population levels vary according to land use. No single data source exists so a composite model is constructed using data from disparate sources. Temporal profiles are derived from surrogate data based on assessments of data sources. A prototype software application is created using a Geographical Information System to enable the distribution of population to be simulated for different times. The intention, until late into the research, was to work with the City of Bradford Metropolitan District Council for a case study but key datasets for this area were not available due to copyright restrictions. The city of Chester in North West England is therefore used as an alternative.
The results confirm the hypothesis that population distribution is concentrated in the city centre, commercial and retail areas, and visitor attractions during the day and disperses to residential areas during the evening. Spatial analyses are able to calculate the likely population within simulated evacuation zones at different times of the day, week and year.
This thesis investigates the current status of the British Columbia Oil Spill Response Plan, discusses the components of oil spill trajectory modelling and analysis, and describes the existing British Columbia Coastal Resource and Oil Spill Response Atlas, which is in essence a simplified Environmental Sensitivity Index (ESI) Map. Two alternative methods are then presented for the creation of multi-factored ESI maps, in the form of raster images that show the relative sensitivity to oiling from an oil spill.
Two processes are demonstrated for creating ESI mapping. The first, an ArcInfo-based Boolean approach shows discontinuous areas of high, medium and low sensitivity to oiling. The second, using IDRISI and Multi-Criteria Evaluation Weighted Linear Combination, defines the continuous degree of sensitivity to oiling over the analysis area.
Finally, an oil spill scenario is generated using GNOME, the General NOAA Oil Modelling Environment, to simulate a spill within British Columbia ’s Strait of Juan de Fuca , using scenario parameters derived from historical marine accident date for the area. The oil spill trajectory model results are then overlain with the ESI map data, an approach referred to as receptor site analysis, utilising an oil probability density surface and volume contours, developed through a quartic approximation of a Gaussian, or Normal , kernel function.
Resource managers use landscape metrics to quantify the spatial complexity of landscape patterns. Subsequently, these metrics supply the means to compare landscapes across space or track changes over time. A complementary method for tracking changes over time is provided by spatially explicit change-detection maps.
Understanding the limitations of thus obtained data fortifies their proper use and sensible interpretation. The current research focuses on the impact spatial distortion has on the comparability of landscape metrics and on change-detection measurements. Spatial distortion was simulated by means of linear polynomial rubber-sheeting either after alteration of the map projection system, or after virtual re-positioning of control points. The former scenario mimics a situation where map projection information is missing for one of the categorical maps included in the analysis, while the latter scenario imitates a situation where one of the sources appears to be un-projected.
The study regarded a number of frequently used spatial configuration metrics at patch-level (patch area, patch perimeter, patch shape index, patch fractal dimension, patch edge-to-edge nearest neighbour distance, and patch centroid-to-centroid nearest neighbour distance) and their class-level counterparts. Spatial distortion was found to have no significant impact on the selected landscape metrics.
However, spatial distortion was found to have a significant impact on change-detection measurements. The amount of tolerable spatial distortion depends largely on the focus of the enquiry and the spatial characteristics of the data set. When the focus is on the entire landscape, and not on individual patches, higher levels of distortion are acceptable. When the mean patch size of a categorical map is larger, higher levels of spatial distortion are tolerable.
A significant relationship has been established at patch-level between the root mean square error ( RMSE), the patch area, the patch perimeter and the patch fractal dimension, and the relative overlap a patch still has with its original counterpart after distortion. RMSE is a measure for overall spatial distortion; while patch area and patch perimeter are both measures for patch size, and patch fractal dimension for patch shape. Change-detection error increases significantly, as spatial distortion increases, as patch size decreases and as shape complexity increases. The multiple linear regression model, linking these variables, explains 46.4 % of the variability that exists in the relative patch overlap data. The model’s predictive capacity has proven not to be enough to facilitate confidence intervals for change-detection overlap figures measured on mis-registered data sources.
Hence, resource managers can rely on the selected landscape metrics’ outcome, even when registration accuracy is doubtful and providing all other sources of error have been eliminated. However, spatially explicit change-detection figures are not always reliable under distortion and can not be predicted without further study.
Top of page
|
