To meet the above challenges, the California Resources Agency Legacy Project entered into an Interagency Agreement with the University of California Davis, Information Center for the Environment. The agreement was charged with collecting all available digital general plan maps from county and local entities in California. Where these digital maps where not available, the agreement provided for creating the digital maps from paper maps provided by the county or local entity. The second phase of the project required the University to build a crosswalk of the source data's land use classification system, which would standardize the language defining land use in these maps.
A significant impact to California's natural resources, human and built infrastructures is the increasing development of rural ranches. Developments of individual houses on 20 to 40 acre parcels has the potential of not just occupying a tremendous amount of California's open space, it might also negatively affect the states farming capacity. Moreover, this kind of development can easily overwhelm the rural transportation infrastructure build for much lower densities and have significant impact on water use and air quality. As such, this agreement requires the University to research and put in place a 'Very Low Density, Rural Residential' land use class. This class requires academic research and expert opinion to articulate where on the landscape this kind of development a) has happened b) is planned, and c) is likely to happen.
The end product of this effort will be two Geographic Information Systems (GIS) digital maps. These maps will be placed in the Resources Agency Legacy Projects California Digital Conservation Atlas, and the California Environmental Research Evaluation System metadata Catalogue. The data will be free to download for anyone to use.
By partnering with the University of California Davis to create this map, the California Resources Agency Legacy Project leveraged expert academic research with the planning landscape. This partnership will allow this data to be released free of charge to anyone who needs it. The fact that this partnership happened means that any of the planning entities at the federal, state or local will not have to re-invent the wheel to acquire the one planning area in which they are interested. Moreover, the map standardizes the complex planning language to unique and useable class of information. The map saves the state money by bringing all of this local data into one map.
First, we will discuss methods that we did not use. A good method would be to use parcel data with a full assessor's table in which most residences will have a homeowner's tax deduction value. The occurance of the residence tells us that it is a residential parcel and the other data give the parcel size. Unfortunately, it is impossible to get parcel data for all counties and so we developed a second method, which could be used on all counties. We developed and validated our method on those counties for which we have parcel data, which include the Central Valley counties from Colusa to Kern .
We inferred the LDR and VLDR areas by using census block population data, which we purchased from Geolytics, Inc. These data contain the 1990 and 2000 population data, with both years in the 2000 block boundaries. Because these two data sets are in the same boundaries, we get change in population between 1990 and 2000 for all blocks in California. We can convert the population density in 2000 to dwelling density, and we can also calculate absolute change and rate of change in dwellings. The LDR and VLDR areas are derived from acres per dwelling unit (acres/du) in 2000. We supplemented this analysis with data on the growth rate from 1990 to 2000, in cases where there was a borderline dwelling density in 2000 and we couldn't decide whether to classify an area as LDR or VLDR or agriculture. If the growth rate was high, we classified the area into the higher of the two categories at issue. We give the detailed steps used in ArcView, below.
The Meaning of The General Plan Data Layer From our data, one can see that, in many counties, much of the area designated Agriculture in the general plan in fact permits rural residences, which we define as LDR (parcels of 2 to 20 acres) and VLDR (parcels of 20-160 acres). For many habitat protection purposes, the VLDR land uses will not interfere much with habitat values. However, for some large mammals, even 20-acre or 40-acre parcels can interfere with their movement, due to noise, dogs, and vehicles on roads. LDR land uses will typically interfere with most vertebrate species. Also, note that our method results in a minimum delineation of LDR and VLDR areas. We have only mapped areas that were occupied in 2000 and not other areas zoned for LDR and VLDR land uses but not yet occupied.
It is also of interest to see that many counties permit LDR over much of the Agricultural general plan area. This means that commercial agriculture will be phased out of these areas in the future. A few counties permit LDR, or more intensive uses, over all of their private lands. Some counties, such as Kern, allow large areas of LDR on farmlands, but keep it all contiguous to existing cities. This method preserves commercial agriculture, as long as possible. Also, of interest is the fact that even VLDR uses can create conflicts with commercial agriculture, especially due to the effects of arial spraying drifting onto rural residences, resulting in the phasing out of agriculture, even when there are only a few residences built in the rural areas. So, even VLDR can conflict with farming.
The following steps were taken in ArcView software to define LDR and VLDR: 1. Select luc = 1 in the county general plan table and convert them into a shapefile, which is named countyname_gpa. This shapefile is all of the agricultural land use designation. The low density residential designations in the county general plan, if any, will be kept as low density residential, as they are not in the agricultural polygons and so are not affected by our calculations. 2. Calculate acre/du for each census block, for 2000, in the census population layer countyname_cenblk00_90. 3. Use countyname_gpa to clip countyname_cenblk00_90 to get acres/du of each census block, but only within the general plan agriculture designation. The new shapefile is named as countyname_clip1. 4. In countyname_clip1, if acre/du < 100, then luc = 7, which is LDR; if acre/du > 100 and < 160, then luc = 13, which is VLDR. In the table of this shapefile, a new field "le_gpdes" which means "Legacy Program general plan land use description" is created so that both the new and original land use designations in general plan can be kept. We use a cutoff of 1 du < 100 acres, rather than the density threshold for our LDR category, because we are calculating density on whole census blocks and so a density of greater than 1 du on 100 acres implies smaller parcels somewhere within the block. We validated this method by examining developed parcel sizes in several counties for which we have parcel data. The developed parcels in these test counties are about 20-acres or smaller in the census blocks when we set the average density threshold for the blocks at 1 du/100 acres. We assume, due to legal and political considerations, that once a significant fraction of parcels are developed at LDR density that nearby parcels will also be able to develop at this density. So, we use the blocks as the initial spatial unit in which to make this calculation. 5. Union countyname_clip1 with countyname_gp, the county general plan layer, and name it countyname_gpt which represents the Legacy program temporary general plan file 6. Combine polygons to create spatial continuity of land use designations in countyname_gpt to create a new general plan which is named as countyname_gple in which le means Legacy Program. We do this judgementally, to create areas that consist of mostly census blocks with an average density > 1 du/100 acres. As these areas represent zoning designations, we included the intervening blocks as LDR, in order to tie together the blocks that met the density threshold. In general, our LDR designations look similar in size, or smaller than, those in the county plans that contain such designations. Also, zoning, to be legal, should generally include many parcels or large areas of parcels, so most counties use fairly big polygons for each category in rural parts of the county. Where county general plans included LDR and VLDR categories, we tested these against the census data, but in most cases simply kept the county designations.
Post Processing Notes The following steps where taken in Arc/Info software to post process that data. 1) The Resources Agency is interested in delivering seamless data for conservation planning and resource investment planning. We would like to have this data set meet the general needs of 'seamlessness', data consistency and data quality. 2) The General Plan layer should have a consistent attribute structure. 3) The Resources Agency added and ensured consistency for County Names, FIPS Codes, and Alpha-numeric (county number) codes. 4) The Resources Agency ensuree a consistent spatial context for county boundaries
The shapefiles delivered had excellent consistency for the LUC coding scheme. However, it had poor consistency for county boundaries. This problem stems from the multiple sources, multiple jurisdiction and use of Tiger files (most likely). Resources Agency went through the following process per county shapefile. 1) Convert to Arc/Info coverage 2) Ensure attribute consistency (item definition and population) 3) Dissolve on LUC 4) Clip to a standard county boundary (Using the Teale County Coverage). This step necessitated removing all General Plan polygons outside the Teale County Coverage, and attributing all new polygons inside the Teale County coverage and outside the General Plan coverage with the General Plan LUC for the adjoining General Plan polygon. A dissolve on LUC then happened again. If we did not run this step, then there would be tens of thousands of sliver polygons along the county edges. 5) Add and populate county name, FIPS and number 6) Assemble statewide coverage per county. 7) Peer review a set of selected counties. 8) The process was performed for the GP data and the LE data.
Peer review was performed by contributing memnbers of the California Planning Roundtable, under the direction of the President of the California Planning Roundtable. Members reviewed data from counties they represent.