Scope and Limitation

No data is completely accurate and historic data is certainly no exception.  We have tried to locate the boundaries of the historic districts and the inventory items in them as accurately as we can but a cursory examination of the maps shows many locational issues.  This document attempts to explain the processes used to locate the boundaries and inventory items.  It wou ld have been great to have the time and resources to visit every district in the state with a Global Positioning System (GPS) unit but that was not possible.  That will have to wait for another time.

Historic District Boundaries

Local/Parcel: For some, but not all of the communities there was an available layer of land parcels (approximate date 2006) supplied by the Connecticut Department of Energy and Environment.  If the original study commission report map had parcels boundaries drawn on it the DEEP parcel layer was used as the data source for the district boundary.  Occasionally there were differences between the two maps and I made my best attempt to be faithful to the study commision map.

Georeferenced:  In communities with no parcel information I used a GIS process called georeferencing where intersection locations were located on the study commission map and also on detailed aerial photography. The study commission map is pulled up to the location of the air photo with a mathematical transformation using the two sets of intersection coordinates.  The district boundaries processed in this way are only approximate.

OPM: The Connecticut Office of Policy and Management published a Plan of Conservation and Development, 2005-2010, and as part of that process a GIS data layer of historic districts was obtained.  The original source of this data is not known.  For a few historic districts where there were no parcels and georeferencing was poor the boundary from this data source was used.  District boundaries processed in this way are poor.

Visual Fit: Where there were no parcels, georeferencing was poor and there was no district in the OPM layer the district was estimated using a visual fit of 2008 aerial photography. District boundaries processed in this way are poor.

Inventory Location Processes

Address-based geocoding was used to intially locate the items in the inventory list.  This process uses interpolation based on address ranges and results in only approximate locations.  In most districts it was possible to use additional information from parcels, where available, the original maps and an estimate of location based on the address.  Not all items in the inventory could be located so there will be fewer markers on the district maps than entries in the inventory list.

In decreasing level of accuracy:

Located to parcels: In towns where we had parcel information and the original map did as well it was usually possible to place the inventory item at least inside the parcel and frequently on top of the structure visible in the parcel.  To do this we matched the Parcel Identification Number or Map/Block/Lot number in the DEEP parcel layer with the ParcelID field in the inventory. These are the most accurately located inventory items.

Located to addresses in parcels: In some towns the original maps had addresses marked on the maps and where there were no parcel Ids in the inventory this process was used.  It is almost as accurate as the process above.

Located to addresses, edited to structures; locations approximate: Where there were no parcels and the aerial photography was adequate the inventory item point was moved to the structure most likely to have that address based on the ranges of addresses and the visible structures.  This process is not as accurate as the one above and if one item is out of place on a street segment adjacent items on that segment will also be off.

Located to addresses: This is the poorest level of accuracy.  Usually the air photo was inadequate to locate structures.  In urban areas the photography was of much higher resolution and taken during tree leaf-off conditions but in rural areas the photography is of much lower resolution and was taken during the summer so structures are not as visible.  This process produces the lowest level of accuracy.