GEO-spatial data accuracy and its legal implications
This paper presents the first results of a feasibility study of a system where photogrammetry, digital maps and GPS are combined to georeference GPR surveys
GIS was first introduced in Malaysia about a decade ago. Amongst the major issues that arose, was the absence or lack of digital geospatial data. GIS users then, needed to digitize their own data from hardcopy maps or survey plans to convert them to digital data. This conversion was certainly time-consuming and expensive, so much so, that accuracy of data was not of paramount importance. At present, with the realization of GIS as a tool capable of storing massive data, manipulating, displaying and querying, modeling, visualization applicable to multiple discipline, has lead to a rapid demand for GIS data and spatial data. The producers and suppliers of these data are no longer restricted to the Department of Survey and Mapping Malaysia as the government agency for mapping, but GIS users can now obtain digital geospatial data from other government agencies such as the Agriculture Department, Local Government Authority, Department of Director-General of Lands and Mine as well as commercial data suppliers and geographic service providers. Consequently, the problems today are caused more by the increasing availability of digital geospatial data rather than the absence or lack of them.
Due to the scores of available data, users are in a dilemma as to which data to access and what is relevant to their needs. Very often, users are simply unaware of the types of data that are in existence and its location. To add to their predicament, with the numerous data providers, the users also have to choose from which acceptable or correct source to access and use, as data may be available from different sources. Disparate data from these sources, though could be integrated, are usually incompatible in terms of storage format, map projection, map scale, symbols, accuracy, and other cartographic specifications. These data therefore need to be evaluated and most often converted. Hence the question of merits of the data came into being.
Users’ perception of GIS data has been positive as they generally believe that it is accurate and reliable. The users of GIS data in Malaysia can be categorized into four broad groups, namely, the experts in GIS, the moderate users, the less expert users, and the non-expert users. The non expert users are often impressed by the beguiling attractiveness and the high aesthetic quality of cartographic products from GIS so much so the accuracy of data is not of much concern. This group of users usually use GIS to display graphics for reports. Although many use GIS, they however, seldom question the accuracy and reliability of this data. Matters such as the map scale in use, the map projection adopted, and above all, the reliability of the base map from which the map is derived were not often broached. While not realizing these factors can affect their decisions, users interpret the information in GIS data according to whatever knowledge and experience they possessed. On the other hand, expert users such as the land surveyors, due to the nature of their profession, emphasis the need for the data to be precise, accurate, relevant and complete. GIS professionals that undergo formal education in GIS both locally and especially from abroad by now know the difference between GI System and GI Science. However many professionals from other disciplines still maintain GIS as a tool for decision making, although some may now know how to use GIS, to do prediction or to model their alternative solutions. Unfortunately these professionals have little or no knowledge of map making and surveying and hence fail to assess the quality of the data used. Moreover, their exposure to GIS is limited to short term training or courses on GIS which are inadequate in content. It can be summarized that Malaysians, except for professionals whose job requires the use of maps, are mostly not accustomed to map reading in their daily activities. Therefore it is understandable that users may not be aware of the accuracy of GIS data.
Geospatial data production
Data Life Cycle
It is observed that changes that occur during the data life cycle (Figure 1) may affect data accuracy. As has been pointed out by Russell G, Congalton H, Todd Morer in their book Quantifying Spatial Uncertainty in Natural Resources : Theory & Application for GIS and Remote Sensing, (2000), common occurrence of changes of data in its life cycle are: during observation;
• during interpretation of observations, e.g. interpretation of vegetation boundaries on air photographs which creates linear entities;
• during digitization e.g. replacement of a smooth, analogue line on a map with a polyline in GIS database
• during resampling associated with projection change or change of spatial resolution: generalization of data; &
• during assembly of results in support of decision or for archiving
In addition to the occurrence of changes during the data life cycle which may affect data quality, the following circumstances show instances when inaccuracy comes about. Geometric incompatibility occur when digital geospatial data captured from different map sheets or obtained from different sources fail to match. This may be due to changes to the geo-referencing standards in which new datum replaces old datum. These problems frequently occur during the transitional period when geospatial data referenced to original datum and new datum are in concurrent use.
Another instance where inaccuracy occurs is during the data collection stage where geospatial data is collected at different scales and with different map projections. National small-scale (1:750,000 and 1:500,000) and medium scale (1:50,000) maps were based on the cylindrical Rectified Skew Orthomorphic projection. On the other hand, large scale cadastral maps (1:5,000) is based on the rectangular grids of the Cassini Soldner projection. Data inaccuracy may also occur even when similar map projection and similar scale were used as the maps are produced by different agencies for different purposes at different times. It requires reconciliation of mismatching features across map boundaries which is clearly not a trivial task requiring human decisions to apply logic to resolve the problem. Another common phenomenon that causes data inaccuracy is the quality degradation due to time. Although data collection and digitizing are carried out using relatively stringent specifications as a rule, the same level of requirements is not always enforced when digital databases are updated. This has resulted in the degradation of geometric accuracy in the contents of the database. It also tends to invalidate the data quality information attached to the metadata of the data sets concerned.
Quality is also degraded when digital databases are not maintained properly. Ideally, a geo-database is a faithful snapshot of the status of human activities and natural features that are found at a particular geographic area of interest at a specific point in time. When these features or activities change, the database must also be updated accordingly. This requires continuous monitoring of human activities and the natural environment in order to check all the changes that have occurred. When such database is not updated, it will lead to serious uncertainties when data is used for time-sensitive spatial problem solving.
Data producers/ providers
In Malaysia, geospatial data producers could be conveniently categorized into two, i.e. government and private sector data producers. Most, if not all of these producers rely on the Department of Survey & Mapping, Malaysia’s (widely known by its Malay acronym JUPEM) cadastral parcel fabric or topographic data as the base data to collect and produce their own geospatial information. At national level, a committee known as the National Malaysian Spatial Data Committee (formerly called the National Mapping Committee) was formed to coordinate the data acquisition activities of these government geospatial data producers.
Related issues of data acquisition and production, including addressing the needs of members and data quality were worked upon by the various subcommittees formed under it. The activities of geospatial data producers in the private sector were however left uncoordinated and not very much controlled except for the conduct of cadastral surveys and production of cadastral survey data. Information on their activities are only known in two circumstances, i.e. when application is made to JUPEM to collect data and when those acquired data were submitted to JUPEM to obtain clearance for geospatial information production. Unlike in the case of cadastral data, vetting of other geospatial data by JUPEM only involves the filtering of security sensitive information, and not quality or more specifically accuracy inspection.
JUPEM remains inarguably the major producer and provider of geospatial data in Malaysia. It retains ownership over those data in the form of copyright over its product. Nevertheless, data ownership has brought about issues of legal liability founded on the “harm-based concept” in which harm or injury incurred as a result of errors or shortcomings or incorrect decision could well be due to data inaccuracy. In such instances, in determining the liability of parties involved in the handling of geographic information, i.e. from the original data providers/ producers, software producers, secondary producers and finally the users, the law will look at those in the information chain and consider whether they have exercised appropriate standard of duty to prevent the occurrence of the damages. The law that is applicable to ascertain liability is the law of contract and tort. However, there is no specific legislation on the matter. JUPEM had in the past provided disclaimers exempting them from liability, on its printed maps and recently, on its digital map products; however, other government agencies had varied practices pertaining to this matter. Although there has not been any known litigation against them, all currently acknowledged the fact that users are getting more sophisticated and increasingly aware of their rights and as such many had taken steps to protect themselves in the eventuality of providing inaccurate data to their clients.
However, the liability in geospatial products and services relating to computerized geographic information systems is difficult to determine by traditional legal theory. This is because of the wide array of current as well as potential application of geographic information technologies. Each application requires integration of information specific to the application and often will involve different attributes, analytical method, spatial features and accuracy requirements. Other legal issues that have to be tackled by JUPEM and other data providers should include identifying duties which are mandatory as data providers regarding the quality of data; duties that every professional is expected to do (Bedard, Devillers, Gervais & Jean-soulin, 2004). Among these duties are the responsibilities of informing users about the datasets, that not only provide users with information pertaining to the content of the data but the limitation or defect or potential risk in the data utilization. In other words, the data producers need also consider users’ intended usage of the data and warn them accordingly. These legal obligations or ethical requisites may be provided under the code of conduct as provided by the Licensed Land Surveyors Act 1958 (Revised 1991) or the consumer protection legislations.
The standards of data accuracy vary from one producer to the other, and may be very significantly different due to the fact that their production serves differing purposes. In the case of JUPEM, their town and city maps are produced at scales ranging from 1:1,250 to 1:10,000; the planimetric accuracy of these maps range respectively from 0.6m to 5.0m. Additionally, topographic maps are produced at scales of 1: 25,000 and 1: 50,000 and their planimetric accuracies are 12.5m and 25.0m respectively. On the other hand, the height accuracies would be at half the contour intervals depicted on those maps.
Another source of geospatial data produced by JUPEM is the cadastral survey database. The database was originally developed through the keying-in of bearing and distance values appearing on certified plans, which were derived from actual ground surveys. These cadastral surveys vary in accuracies as they were performed previously under three different categories or classes, according to the needs of meeting the required level of accuracies; for instance surveys in town areas would need to achieve higher than 10cm level of accuracy, whereas in the countryside accuracy of 50cm is considered adequate. Currently, geospatial data produced from the conduct of cadastral surveys evidently were the most accurate and as such were relied upon to serve as base data for geospatial data production by others. With the most dependable accuracy attribute, it also served as one of the core datasets of the nation’s spatial data infrastructure.
Beginning 2006, the development of the National Utility Database was started by JUPEM, whereby data on the location of underground utilities such as gas, water and sewage pipelines as well as telecommunication and electric cables were captured and stored in the said database. Those data were initially sourced from the utility providers and due to the differing reliability of the information provided, they had to be segregated into four quality levels, i.e. Quality Levels (QL) A to D, with QL A being the highest level in terms of accuracy, with ±10 cm planimetric and height (depth) accuracies.
Other governmental producers of geospatial data such as the Departments of Agriculture, Mineral and Geoscience, Planning, Forestry etc., have their own accuracy criteria to meet their needs and specifications but is clearly not as demanding as that of JUPEM.
Apparently, the need for high accuracy geospatial data is most evident in the case of underground utility data production and use. It would be obvious that erroneous data can lead to erratic digging in the course of emplacing new facilities and this could further cause accidents resulting in extensive damages, including the loss of lives. The need for highly accurate cadastral surveys too has been recognized as being imperative and given a lot of emphasis in the past. As such, concerned surveys have been traditionally very tightly regulated and this has resulted in a reliable cadastral survey system that underpins the highly progressive land market of Malaysia.
Geospatial data accuracy specifications or statements were in the past not given much attention. Nevertheless, over the last decade or so, demands have been made by users for data producers to publish quality or at least accuracy statements for their data. JUPEM has responded to this call by publishing data accuracy statements in the metadata published through their on-line JUPEM Geoportal (a dedicated departmental website to provide on-line sales of data and services to users).
Efforts to standardize the measure of data quality (including data accuracy) and publishing them had been initiated over the last few years. JUPEM has taken the lead in this effort, whereby the task of determining data quality through field measurements and verification had been conducted for the map sheets that they had produced. This form of verification, albeit laborious, is deemed necessary and the outcome of this quality check is published in the metadata produced by the department. JUPEM had also been engaged in developing the data quality standards which would eventually be utilized by all geospatial data producers.
Legal implication from data inaccuracy
Malaysian is a non-litigious society. There is no known legal suit pertaining to injuries or damages arising from data inaccuracy. However, court cases with regard to damages suffered as a result of data inaccuracy, in developed countries where GIS originated are on the rise. Issues of liability as a result of loss of earnings, opportunities, property and even life gave rise to questions of ownership or authorship which have become more uncertain as data can be easily manipulated and mixed with data from other sources, sometimes of unknown lineage and perhaps at inappropriate scale. Liability has been shown to possibly arise from inaccurate, incomplete and misleading information of data as well as incorrect decisions. The first indications that there are serious problems with the data are when accidents take place (Cho, 2005).
It has to be noted that on the international front, Malaysia had on two occasions appeared in the International Court of Justice in disputes with her neighbors, Indonesia, over the islands of Ligitan and Sipadan, and with Singapore, over Pedra Branca. Amongst a multitude of factors put forth in the arguments, aspect of data inaccuracies was also hinged upon in both of the aforementioned cases.
Since there are no court cases from Malaysia to illustrate the damages or injury that occur owing to data inaccuracy, the following cases from USA and Australia are referred to indicate the legal implication as a result of error in maps, that can be fatal:
According to a report in the US journal Point of Beginning as cited by the Asian Surveying and Mapping (11 March 2009), mistakes by surveyors in Texas are being blamed for millions of dollar losses suffered by local landholders as a result of flooding in the aftermath of Hurricane Ike, the third most destructive hurricane in the USA. Some of the damage in the US was due to flooding, where housing had been built below the Base Flood Level. This is a contour defined by the Federal Emergency Management Agency to correspond to a 100 year flood level. The position of the Base Flood Level was fixed by the National Geodetic Survey long ago and marked by concrete and brass monuments. In the 1980’s the Federal Emergency Management Agency (FEMA) re-measured the contour and found that the flood plain was about a meter above the old marks. It issued a new map that become the document of reference for insurance companies and other authorities. However, the surveyors continued to rely on the old elevations. As a result of the mistake made by the surveyors, 20 homes near the town of LaBelle were built in the flood plain as people who thought their new homes were being built above flood level were actually building a meter below it. To add to the homeowners’ nightmare, they had no flood insurance and FEMA would not permit them to rebuild their homes because they were in the flood plain.
Four New South Wales, National Parks & Wildlife Service (NPWS) officers were killed from smoke suffocation in a burnoff operation. The officers were given maps that showed two possible escape routes, but which ended in impenetrable bush or line of cliffs. The Senior Deputy State Coroner informed the court that there were deficiencies in the maps used in the operations as information contained in the map issued by NPWS showed a cleared hilltop, which potentially could have provided shelter from the fire. Unfortunately, the map did not show a 30m cliff which stood between anyone trying to escape the fire and the cleared area. The map also showed a path known as Wallaby Track running directly towards a local motorway. In reality this path twists into impenetrable bush. It was found by the court that the original botanical map had not been ground-truthed to include specific details and did not mark areas with safe refuges to retreat to as required in the fire management procedures guidelines (Cho, 2005).
The above two cases demonstrate the classic issues of standards to be attained if liability is to be avoided. The duty to take care, the responsibility of due diligence to those who may be affected by a lack of care, the reliance on information to one’s detriment, and the subsequent injury, damage and loss that occurs, are established legal standards. Such standards are set either by statutory mandates or through the common law. In Malaysia, however, there’s no statutory mandate in order to ensure compliance. However, the common law of tort may be applicable in failure to meet the required standard of care by the profession.
The rapid growth of GIS users in Malaysia has raised the need for data accuracy and the appropriate management of these data. There is no specific law concerning data quality, and no consistent legal framework on the management of GIS or geospatial data; what exists is simply a patchwork of self regulation in the form of government circulars and statutes. It is thus important that Malaysia have a codified policy on managing geospatial data as the country moves towards a spatially enabled government. This will enable relevant bodies to be given statutory mandate to ensure effective and up to date collection of data and the imposition of standards to be followed by data producers so as to ensure data quality. It is also important for the government and institutions of higher learning to give adequate emphasis in formulating appropriate programs to create awareness on the importance of data accuracy.
Abdulharis, R., B.van Loenen & J.Zevenbergen,(2005), “Legal Aspects of Access to Geo-Information within Indonesian Spatial Data Infrastructure” in ISPRS Workshop on Service & Application of Spatial Data Infrastructure, XXXVI, Oct. 14-16, 2005, Hangzhou, China at pp.147-153
Bishr, M., Wytzisk, A., Morales, J., (2007) “GeoDRM: Towards Digital Management of Intellectual Property Rights for Spatial Data Infrastructures” in Research and Theory in Advancing Spatial Data Infrastructure Concepts, ed. Onsrud, H, Redlands, CA: ESRI Press
Bedard, Y, Devillers, R, Gervais, M, & Jean-soulin, (2004) “Towards Multidemensional User Manuals for Geospatial Datasets: Legal Issues and their consideration into the Design of a Technological Solution” of the Third International Symposium on Spatial Data Quality (ISSDQ’04), 15 -17 April Bruck an der Leitha, Austria
Blackmore. M and Longhorn R. (2004) “Ethics & GIS: The Practitioner’s Dilemma”, in AGI 2004 Conference Workshop on “GIS ethics”, 14 October 2004 London, England, UK
Cho, G., (1998), Geographic Information Systems & the Law Mapping the legal frontiers, John Wiley & Sons Ltd, West Sussex, England.
Cho, G., (2005), Geographic Information Science Mastering the Legal Issues (2005),John Wiley & Sons Ltd. England.
Cho, G, (2007) “National Spatial Data Infrastructure, intellectual property rights and geospatial technologies in aiding economic growth” in 6th Annual International Conference and Exhibition on Geographical Information Technology and Applications, Kuala Lumpur, Malaysia.
Mohamed, A.M., (1998), Case Studies of NSDIs in countries in transition – Malaysia, October 1998 as assessed via internet on 8 February 2008 in http:// gsdidocs.org/docs1998/canberra/malaysia.html
Mark R. Leipnik, Donald P. Albert, GIS in Law Enforcement, 2002
Nordin, A.F., (2007), “Spatially Enabled Government : The Malaysian Case” in the International Workshop on Spatial Enablement of Government and NSDI Policy Implications”, 12 Jun 2007, Seoul, Korea.
Onsrud, H.J. (2004), Geographic Information Legal Issues, in Encyclopedia of Life Support System (EOLSS), Developed under the auspices of the UNESCO, EOLSS Publishers, Oxford, UK
Onsrud, H.J. (1999), “Liability in the use of GIS & Geographical Datasets” in Longley, P., Goodchild, M, Maguire, D and Rhind, D (eds.), Geographical Information Systems :vol 2 Management Issues and Applications, John Wiley & sons, pp 643-652
Onsrud, Adler, Archer, Licensing Geographic Data & Services, 2004 National Academies
Saxby, S. Dr. , “Public Policy and the Development of a UK National Geographic
Information Strategy”, 2006,International Journal of Law and Information Technology Vol. 14 No. 2 Oxford University Press
Gehan Gunasekara, The ‘final’ privacy frontier? Regulating trans-border data flows, 2006,International Journal of Law and Information Technology International Journal of Law and Information Technology Vol. 15 No. 3 Oxford University Press
Jane Holder, Carolyn Harrison, Law and Geography, 2003, Oxford University Press, London
Ida Madieha Azmi, “Personal Data Protection : The Malaysian Experience”, in Information & Communications Technology Law, June 2007, Vol.16, Issue 2 pp. 125-135 as assessed on 20 August 2009 from http://web.ebscohost. com/ehost/delivery?vid=25&hid=0a4335b1-57de-4fal
This was an invited paper at the Eighteenth United Nation Regional Cartographic Conference for Asia and the Pacific , Bangkok, 26-30 October 2009.
INDUSTRY | GPS | GALILEO UPDATE | LBS | GIS | REMOTE SENSING
Mark your calendar