A GIS Workshop Project by:


Jack L. Lyle


University of Texas at Dallas


Summer, 1998

Email: [email protected]

Project Objective

Project Methodology

Project Construction

Project Results

Project Conclusions









Although the analysis of land titles and the issuance of title insurance are inherently geographical, neither I nor any real estate professionals I have queried are aware of any use of GIS in the land title industry. County tax appraisal districts across the country have built large, successful GIS applications to render taxes on real property, and city governments are using GIS to manage their municipal enterprises, including the construction and maintenance of cadastral coverages. So, it would seem logical that land title companies, by adding historical data to already existing cadastral coverages, could use GIS in their activities, and the purpose of this project is to illustrate the feasibility of using GIS in such a manner. RETURN TO TOP





To narrow the scope of work for this project to something achievable in the short timeframe, I decided to limit my activities to two abstracts in Dallas County and to limit the timeframe to the early part of the land history of Dallas County. I selected the John Hyde Survey, Abstract No. 547, and the David Murdock Survey, Abstract Number 996, adjacent abstracts located at the north end of present day White Rock Lake. Research would be done at the Dallas County Courthouse to secure the conveyances in these abstracts from the formation of Dallas County in 1845 to just prior to the construction of White Rock Lake in about 1908.


This time period was selected for two reasons. First, there are far fewer conveyances during the early years of the county than during more recent times, making for a more manageable amount of data. Second, this time period is probably the least documented in terms of graphic data -- there are no maps -- and, therefore, if land title history for this time period can be dealt with efficiently in GIS, then dealing with later periods with graphic data should be achievable, even though the data volume is greater.


These conveyances would be built into an historical GIS coverage/theme. Subsequently, a modern parcel of interest, a "Subject Parcel", could be intersected with the historical parcels and the resultant selected parcels sorted by date to show the chain of historical owners underlying the "subject parcel". This simple GIS analysis and the presentation of the results are the fruit of this exercise. RETURN TO TOP




While performing the courthouse research, at first I was very faithful in following the chain of title. Then I realized that since my study was to determine if GIS was a useful tool for land title applications, I could use the GIS to identify gaps in title, so I became less rigorous in my organization and began to take a scattershot approach, letting instruments guide me to adjoining parcels and to interesting sidelights. I found in building the GIS that gaps in title and in locatability were easy to discover.


Since there are no old maps available, I had to create my own. The instruments had to be harmonized one to another in one geographic scheme, but these old conveyances are on several different directional bearing schemes and three different sets of units are used: varas, chains, and feet. Since I had the North Texas GIS Consortium Street Centerline files, and since the City of Dallas has the old survey lines referenced to modern street centerlines, I chose to use the Consortium projection that is NAD83 with feet units.


I was able to determine the coordinates for a common corner of the two surveys by getting the coordinates of the intersection of Mockingbird Lane and Fisher Road from the Street Centerline files. The bearing for this common line was determined by getting coordinates for another point on the Mockingbird Lane centerline about 3000 feet northeast of the first coordinate pair. These two sets of coordinates were verified by GPS using NGS Station DORAN on Flagpole Hill for control. The Consortium coordinates are within about two meters of my actual GPS positions.


With beginning coordinates and a bearing to use to control the rotation of the historical bearings in the deeds, I was able to begin coordinate geometry (COGO) work to get all the eventual polygons on the same projection. There are at least three different ways to do this, but the work requires someone with specialized knowledge of land surveying and real estate law to order the construction of the polygons; I experimented with all three.


The first way was to use the coordinate geometry commands in ArcEdit to build the polygons and populate the associated INFO files. However, I soon discovered that since I was starting without much of a clue about how the parcels fit together, maybe it would be better to work out all the snags before I started building coverages.


The second way was to do the COGO work in COGO software and import the resulting points and then draw polygons and populate the associated database. This method would work using either Arcview or Arc/Info. I experimented with this method successfully in both packages.


The third way would be to do the COGO work and draw the polygons in a CAD program, then importing the drawn polygons into either Arcview or Arc/Info. I used all three methods, but the majority of work was done using the third method, and supplementing the work with the other two methods.


After all the COGO work was done, I imported the resultant x,y points into a Microstation DGN file set up for data of the magnitude of the NAD83 coordinates. Using the x,y points, I did a connect-the-dots exercise to form the outlines of the historic parcels.


I created polygons for all the various parcels as the lands were sold off. For example, one original 1280 acre survey was sold in two 640 acre tracts. One 640 acre survey was then sold in two 320 acre tracts. The DGN file contained six shapes representing these parcels. The DGN file was exported as a DXF file, which was then imported into an Arc/Info coverage, CLEAN and BUILD were run to cleanup and create polygon topology, and, in the example given above, I ended up with three polygons: a 640 acre polygon and two 320 acre polygons. In building the topology, only polygons of x and y dimensions were built, not the three dimensional x,y, and t (time) polygons I had in mind. This situation caused some work joining polygons and creating duplicate and triplicate polygons for successive sales of the same parcel of land. The polygon joins were accomplished in Arc/Info, and the coverage was then further edited in Arcview to create the sometimes-repetitive polygons that represented different conveyances of the same property. I came to call these polygons poly-polygons. An exhibit showing the historical parcels poly-polygons is attached.


Analysis was then done to check this data, which reside in a single theme called Historical Parcels, to ensure that each conveyance was represented by a polygon and that a complete chain of title existed for each of the tracts of land. This analysis was accomplished by simple selection of a parcel, which in turn selects the underlying polygons, sorting the selected polygons by date, and checking for an uninterrupted chain of ownership.


In this whole process, I found what I call "renegade polygons". These are polygons which were built during the Arc/Info CLEAN and BUILD process which are the quantification of gaps or overlaps between parcels which were thought to adjoin. This turned out to be a great way to check for such title problems. An exhibit showing some renegade polygons is attached.


A theme was created for modern streets by downloading the NTGIS Consortium Dallas County Street File and editing them down to just my project area. Then a surrogate theme to represent a modern cadastre was created and named Subject Parcel. Parcel polygons were drawn in this theme to represent modern ownerships on which a chain of title was to be run. RETURN TO TOP




The process for running a chain of title is very simple. In the Subject Parcel theme, the parcel to be analyzed is selected. Then, the Historical Parcels theme is made active and from the THEME menu, Select By Theme is executed. The theme table is opened, the table is sorted by date, and, from the THEME menu, Promote is executed. All the data for the geographically underlying polygons are now in chronological order and at the top of the table. Convert to Shapefile is executed from the THEME menu, and the chain of title becomes a theme from which to create a report or layout. Two sample layouts, one showing a complete chain of title and the other an incomplete chain are attached.






GIS is a viable tool not only for doing land title analysis but also for building the data to be analyzed. The ability to query and draw the results of the query makes the "debugging" of data very easy. If instead of old handwritten deeds, graphic information were available to be scanned, the construction of the historical database would be even easier.


With many tax appraisal districts using GIS and with that current data available at low cost as public records, the major cost would be in building the historical background. With many title searches only going back twenty-five years, that much data would be relatively easy to accumulate and build. In addition, the ability to deal with easements, tax liens, and other matters spatially should make deciding their effect on an ownership much easier. But, perhaps the greatest impact would be the need to build the database only once instead of doing it piecemeal and repetitively for every title analysis performed. RETURN TO TOP