The three prime entities that are involved in the maintenance of Land information system (LIS) are the Department of Revenue, Department of Survey, Settlement and Land Records, and Offi ce of land registration. In addition to this, the appropriate municipal or village administrative unit is also involved. This is the convention in many states in India. The land records in India, do not provide ultimate title to property. In other words we need a system of effi cient maintenance of land records which is essential for protection of land rights of several
millions of people in the country. Land has been one of the most sought
after possessions in developing countries such as ours, wherein, the land holding has socio economic significance and status. Thus there is a need to develop an integrated LIS. A lot can be said and discussed about various activities and issues that are being dealt to create a LIS in different parts of the country. There are some success stories in Karnataka and some efforts of Governments in the states of Andhra Pradesh, Gujarat and Madhya Pradesh. The importance has been recognized and limitations of the current system are understood at various levels. . In all these success stories there is a systematic implementation and addressing of local issues and sentiments. As such development of an integrated LIS cannot be viewed as just putting all the information in digital form and designing a front end or a typical GUI for user interaction by availing services of few IT experts or MNCs. In fact the Land management in our country is purely Indian way and it has legacies of history and no imitation or service of prime MNCs or international bodies has relevance or use or signifi cance for our country. The basic issues related to data particularly the cadastral maps in terms of micro level components, macro level components, logical consistency and lineage need to be addressed in detail. The technology for land management should be professionally precise following the principles of surveying. Ideally it should also be socially acceptable and economically viable.

Out of several issues an updated cadastral map, graphical information of the parcel with exact co-ordinates of the corners of the fi eld and extent of the fi eld and attribute information about land use, soil type, details of crops and ownership are the most sought after details of any LIS. The scale of the map, creation of ground control points, density of parcels and topography of the area are the key factors to be considered in relating the cadastral map and LIS development. The identifi cation and establishment / measurement of ground control points is essential for the accurate creation or maintenance of cadastral maps. There are several case studies involving the conventional ground surveying methods, conjunctive utilization of conventional surveying tools and GPS, aerial photography and satellite imagery. These are in the form of operational applications, technique development programs as well as research projects. One of the essential requirements of the cadastral survey is the satisfaction of the property owner. Hence the re-survey methodology must provide for the faster handling and disposal of the disputed cases, together with the citizen satisfaction. In the context of requirements of the products / information, detailed assessment of the tasks / tools involved in cadastral mapping is to be carried out. A partial assessment is carried out in this paper which addresses only one or two issues related to updating of cadastral map and geo referencing. These are related to role of high resolution data for cadastral map updating and utilization of GPS /DGPS data in post processing and real time processing modes.

Role of high resolution satellite imagery

The very high resolution data from IKONOS, CARTOSAT and QUICKBIRD is available with resolutions of 1 meter, 2.5 meters and 0.60 meters respectively. What are the implications of resolution for mapping and the accuracy? A satellite image with about 2.5 meters resolution is found to have planimetric accuracy better than 6.5 meters after precision processing based on a systematic study and accuracy estimation. In the same study the one meter data is found to provide 3 meters accuracy. As such the direct applicability or relevance of satellite data for cadastral level mapping need to be understood properly. Here, there are in fact three parameters that are found to be relevant and need to be understood. The first parameter is the resolution of the sensor which is understood by every one as it is given under the specifi cations of the sensors. The second parameter is related to the ground resolved distance which is the size of the smallest object detected on the image. For example the LISS III data with 23 meters resolution shows the railway track. Does this mean that the objects of size 1 to 2 meters can be mapped using LISS III data. It is possible only when we are interested in identifi cation of certain long thin features that exhibit high contrast with their background. The third parameter is related to planimetric accuracy. This is concerned with mapping the features in two dimensional domain and it is found to be a combined result of spectral and spatial discernability. Two different sensors operating in VNIR region each with 80 meters or 23 meters spatial resolution but with different spectral bandwidth and channel allocations are found to provide different accuracies. It is found that the spectral discernability also plays a significant role in delineating the features and thus infl uencing the mapping accuracy. The significance of contrast as an infl uence on spatial resolution illustrates the interrelationships between the various forms of resolution and emphasizes the reality that no single element of system resolution can be considered in isolation from others. This is where the infl uence of mixed pixels comes into accuracy delineation. It is interesting to examine the relationships between the number of mixed pixels in a given scene and the spatial resolution of the sensor. It is found that the percentage of mixed pixels is low for high spatial resolution images. The shape of an object feature is signifi cant. Aspect ratio is ratio of length to width. For long thin features like roads, railway tracks etc the aspect ratio is a very high non dimensional parameter. These features are seen on satellite images even though the resolution is more than the size of the feature. As such it can be concluded at this stage that a careful evaluation of different tools and data sources is required for taking up any step related to enhancing the value of the existing cadastral maps which are on 1 to 4000 or 8000 scale. We should keep in mind the accuracies that are feasible and level of details that are to be identifi ed for land record maintenance. Obviously it should be stated that the high resolution remote sensing data available as on today from different sensors has potential only for thematic mapping and limited utility for cadastral mapping or updating. Let us in all fairness do not hesitate to conclude that high resolution satellite data is not useful in planning, design and implementation of LIS. The only alternative is to go for aerial photography or ALT survey on scales between 1: 6000 to 1: 10,000. Aerial photos where ever available can provide 1 to 4000 scale cadastral maps. The accuracy here is about 40 cm considering 20 microns on the aerial photo negative with 1 to 10,000 scale. Areas requiring survey or resurvey need to go for this as no other method can substitute these efforts to provide the desired accuracies for delineating land parcels The usage of the fi nal cadastral maps using the photographic images or the satellite images is also very limited as the legal boundaries of holdings cannot be determined from the photographs without extensive checking on the ground. In all cadastral surveys there is a need for follow-up ground surveys to check the actual location of legal boundaries that may not be visible on the photographs or may have been wrongly identifi ed. One of the great disadvantages of aerial or satellite survey is that, except for the ground control marks that are part of the basic control network or framework, it does not leave permanently demarcated points on the ground.

The second aspect is related to the application of GPS in differential mode. For areas with reasonable cadastral map availability, the major task of maintaining land records and creation of Land information system can be based on DGPS surveys which can provide scope for georeferencing of the cadastral maps This is possible in post processing mode or real time operation mode. The effort required for post processing mode is more than 10 times the effort required for real time processing mode. It is essential to fi x the most important feature namely the “control point” i.e., bi or tri-junctions (point where boundaries of villages meet) as well as other identifi able points. Presently cadastral maps do not contain information about the height. As such height information can also be incorporated in DGPS stations. This will lead to 3d cadastral mapping in due course of time. DGPS can provide accuracies better than 50 cm to the control points. This is adequate for cadastral surveys. On a map of 1 to 4000 cadastral map the smallest feature that can be mapped is of 80 cm or say 1 meter size. So the accuracy is far more than the minimum mappable unit on 1 to 4000 scale which is in fact
the major satisfying feature. There is a need to create local reference frames or grids which can be used for GPS survey by triangulation or grid methods. It may be noted that the fi rst and second order points built under the CARTOSAT utilization projects by Department of Space, Government of India, all over the country are very useful for the purpose of establishing control points or control grid and traverse running. Access to this information is limited and possibly requires approval and clearance by Government.

This is essential task of the entire ILIS and basic issue. How to go about it. With out much introduction and explanation to the problem it can be stated based on studies in different situations, DGPS in real time operation mode is a better option or solution for this problem. It is not going to provide the solution on a turn key mode. This has only scope and need to be calibrated. This will help us to draw a road map for digital cadastral map creation with minimum of the resurvey task. The cadastral record information with reference to parcel numbers is to be identified with the help of land records and related government offi cers. This is a very crucial stage of the creation of ILIS. How to geo reference the parcel
boundaries. Also these are to be put under a standard reference system like WGS 84 system. It is the major issue. Several techniques of updating / resurveying are being proposed. These include aerial triangulation, use of total station and GPS/DGPS, satellite image processing etc. A detailed analysis and understanding of the potential of the technology suggest that a RTK GPS (It is a DGPS with real time processing capabilities) based surveying of selected points or boundary points and calibration is “the technological tool” for this task in the present context. The experience with RTK was successful in retrieving the control points and the method was accurate for planimetric applications and is several times faster than traditional methods. Generally there are some conditions which to become aware of to obtain a satisfying accuracy by using RTK. There must be contact to a good number of satellites and PDOP has to be low which indicates good satellite geometry. To apply the RTK method in cadastral surveys it is necessary that the basic point has a high horizontal accuracy, whereas it is suffi cient for the determination of the z-coordinate to use a height from a topographical map (the accuracy of the contour lines is about 1 m in topographical maps of scale 1:25.000). A brief account of the achievable accuracies using DGPS is given in Table 1. It is possible to use the GPS method RTK for cadastral surveys. The accuracy of the parcel coordinates is proportional to the number of ground control points used for calibration of GPS and georeferencing. The value of the cadastral maps in such a situation will be enhanced signifi cantly, because every measurement on the map is in a referred coordinate system. In conclusion it can be stated that the Cadastral map based LIS must follow also a standardized data format and prepare a protocol of accuracies so that uniformity is maintained all over the country.