This is a version of the NAWQA (National Water Quality Assessment) "enhanced" National Land Cover Data (NLCDE) that has been further revised using 1990 and 2000 population density by block group to indicate recent (2000) urban development.
NLCDE was developed by NAWQA National Synthesis teams using ancillary information from USGS historical land use and land cover ("GIRAS") data to modify 1992 National Land Cover Data (NLCD).
This data set (v. 3) replaces the previous version (v. 2) dated September 2005. Version 2 of this data set considered any area having 2000 population density of 1,000 people/mi2 as being recent urban and did not consider that the area could already have been urbanized in 1990. The previous approach was determined to have misclassified lands that already were urban in 1990 as newly urbanized.
This data set consists of 29 1-km resolution land use/land cover grids: 28 "percentage" grids and 1 "majority" grid. These grids were generated from NAWQA's version of the 30-meter National Land Cover Dataset enhanced with GIRAS and then revised with 2000 population, or "NLCDEP".
There are 21 land cover classifications in the original NLCD.
NLCDE has 4 additional classes: GIRAS residential (but not NLCD water, urban or forest), GIRAS residential (and NLCD forest), GIRAS orchards & vineyards, and GIRAS tundra.
NLCDEP has the 4 GIRAS/NLCDE classes plus: urban 2000 without trees and urban 2000 with trees.
The final land use class is "missing data."
A 1-km percentage land cover grid was created for each of the 28 land cover classifications. The cell value for these grids is the percentage of 30-m resolution grid cells in the 1-by-1 km area with a specific land cover classification. The percentage land cover in each 1 square kilometer cell was determined by first resampling the 30-meter resolution NLCDE into 40-meter cells, then determining its percentage composition. Each percentage grid is named nlcde_## where ## is the 2-digit numeric code for a land use/cover classification. (See supplemental information for land cover codes and classes.)
A 1-km majority land cover grid (named "nlcde_clsg") was created by determining the land cover class having the highest percentage in each 1-by-1 km area and assigning the land cover code having the "majority" class to the majority land cover grid.
This data set primarily serves to display nationwide land cover but can also be used to compute ancillary information for large study areas where geoprocessing at the 1-km resolution would suffice.
NAWQA needs land use data representing 2000 to indicate land use changes since the 1990's and also to allocate nutrient loads to certain land use categories for water quality trends work. Because 2002 NLCD is not scheduled to be available nationwide until 2006, the short-term solution is to revise NLCDE with 2000 population to indicate new urban development.
This data set is designed as a stopgap measure to characterize 2000 land cover until the 2002 NLCD is available.
The rationale is:
1. Most arable land in the U.S. already is cultivated, so we assume a negligible increase in agricultural land from 1992-2000.
2. The largest category of land use change would be conversion of agricultural land use to suburban and urban land use.
3. The increase in suburban-urban land use can be approximated by the increase in population from 1990-2000.
The data is stored in 28 percentage grids and 1 majority grid. All files are in one directory called nlcdep1km.
To extract the data directory, use WINZIP or the following commands:
gunzip nlcdep1km.tar.gz tar xvof nlcdep1km.tarThis dataset was created using ArcInfo Workstation GRID to update some cells of NLCDE using 1990 and 2000 population density by Census block group. Using population by block groups (instead of the finer resolution block level geographic units) tends to overrepresent urbanized areas, but source grids of population by blocks is not available at this time.
Existing NLCDE water, urban, and wetlands have been preserved.
Recently urbanized areas are those that were NOT considered "urban" in 1990 but ARE considered "urban" in 2000 according to the population density cut off of 386 people/km2 (1,000 people/mi2). The previous edition of this data set considered any area having 2000 population density of 1,000 people/mi2 as being recent urban. The previous approach was determined to have misclassified lands that already were urban in 1990 as newly urbanized. The current version (v. 3) has resulted in a smaller area being called new urban land than the previous version.
If 2000 population density is equal to or greater than 1,000 people/sq mile, and if 1990 population density is less than 1,000 people/sq mile, and if NLCDE is not any of these categories,
11 Open Water 21 Low-Intensity Residential 22 High-Intensity Residential 23 Commercial/Industrial/Transportation 25 Residential without trees (GIRAS) 26 Forested Residential (Res-GIRAS, For-NLCD) 85 Urban/Recreational Grasses 91 Woody Wetlands 92 Emergent Herbaceous Wetlands 41 Deciduous Forest 42 Evergreen Forest 43 Mixed Forestthen the area has been re-coded as "urban 2000 without trees" (27).
If 2000 population density is equal to or greater than 1,000 people/sq mile, and if 1990 population density is less than 1,000 people/sq mile, and if NLCDE is one of these categories,
41 Deciduous Forest 42 Evergreen Forest 43 Mixed Forestthen the area has been re-coded as "urban 2000 with trees" (28).
If 2000 population density is less than 1,000 people/sq mile (area does not meet "urban" criteria for 2000) or if 1990 population density is equal to or greater than 1,000 people/sq mi (area already was urban by 2000), the NLCDE category has not been changed.
NLCDEP CLASSES (modified from NLCD classification)
* = new NLCDEP class, added by overlaying NLCDE and 2000 population data.
+ = NLCDE class, added by overlaying NLCD and GIRAS data.
All other classes are standard NLCD categories.
VALUE LAND COVER
----- -------------------------------------------------------
11 Open Water
12 Perennial Ice/Snow
21 Low Intensity Residential
22 High Intensity Residential
23 Commercial/Industrial/Transportation
+ 25 GIRAS residential (but not NLCD water, urban, or forest)
+ 26 GIRAS residential (and NLCD forest)
* 27 Urban 2000 without trees
* 28 Urban 2000 with trees
31 Bare Rock/Sand/Clay
32 Quarries/Strip Mines/Gravel Pits
33 Transitional
41 Deciduous Forest
42 Evergreen Forest
43 Mixed Forest
51 Shrubland
61 Orchards/Vineyards/Other
+ 62 Orchards/Vinyards (burned from GIRAS)
71 Grasslands/Herbaceous
+ 72 Tundra (burned from GIRAS)
81 Pasture/Hay
82 Row Crops
83 Small Grains
84 Fallow
85 Urban/Recreational Grasses
91 Woody Wetlands
92 Emergent Herbaceous Wetlands
99 No data
Any use of trade, product, or firm names is for descriptive
purposes only and does not imply endorsement by the
U.S. Government.
Although this Federal Geographic Data Committee-compliant metadata file is intended to document the data set in nonproprietary form, as well as in ArcInfo format, this metadata file may include some ArcInfo-specific terminology.
<http://www.asprs.org/publications/pers/2001journal/june/highlight.html>
The original NLCD land cover data does not portray some land-cover categories very well:
1) Orchards are underrepresented, especially in the east. This is apparent when comparing NLCD and 1992 Census of Agriculture county totals.
2) Alpine tundra is not represented at all in NLCD.
3) Transportation is lumped with commercial and industrial or is not captured at all in NLCD.
4) Residential areas (especially mature residential areas) are often misclassified as forest, grassland or agriculture in NLCD.
Ancillary data layers from GIRAS were used to modify NLCD by burning in certain land cover data from GIRAS to the NLCD land cover map. We decided against using TIGER roads data because of coding errors in the TIGER attribute data (August 2003).
GIRAS data provided:
Orchards/Vineyards Tundra (alpine areas) ResidentialSpecific rules were followed to preserve as much information as possible in the result.
CHANGE NLCD TO: WHEN: PRESERVE NLCD WHEN NLCD IS:
Orchard GIRAS is Orchard Water, Urban,
and state is not Orchard, Urban Grasses
CA, FL, OR, WA
Tundra GIRAS is Tundra and Any category except
NLCD is Shrubland Shrubland and
or Herb. Grassland Herbaceous Grassland
Residential w/ forest GIRAS is Residential Water, Urban,
and NLCD is forest Urban Grasses, Wetlands
Residential w/o forest GIRAS is Residential Water, Urban,
and NLCD is not forest Urban Grasses, Wetlands
The September 2005 version of NLCDe 92 was clipped to the
U.S. boundary and shoreline.
<http://landcover.usgs.gov/natllandcover.asp>
The method to generate individual land use percentage grids and the majority grid was developed by Curtis Price and had been modified by Naomi Nakagaki. The two major processes are to create the percent grids and then to create a majority grid. The two major processes have been set up in AML routines.
/* Received from Naomi 7/14/2005 to create 1km grids /* from NLCDEP (NLCDE REVISED WITH 2000 POPULATION) tiles /* /* PURPOSE: this is a copy of Curtis's agg.aml used to /* create 1km pct grids from NLCD 30m grids (4 tiles) /* I am running it on the [NLCDEP] 30m tiles updated /* with 2000 population. /* /* /* [NN] DATE: 5/18/01 /* NAME OF AML: cr8pctgrds.aml /*kjh March 8 2006 &echo &bri &pt &on /* create 1km pct nlcdep grids from 30m tiles &call cr81kmpctg /* create 1km majority grid from 30m tiles &call cr81kmmajg &echo &off &pt &off &ret /******************************************************************** &routine cr81kmpctg /* create a snap grid for 1 km grid grid verify off setwindow 0 0 1000 1000 setcell 1000 xsnapg = 1 /* 1 x 1 grid with value 1 reset /* do the following for each of the 4 tiles &do tile := 1 &to 4 /* Calculate # 40m land use cells in each 1km cell /* the setcell 40 resamples 30 to 40, which is then aggregated /* to 1000m (25x40) by the aggregate function setcell 40 setwindow nlcdep%tile% xsnapg &if ^ [exists nlcdep%tile%_c -grid] &then ~ nlcdep%tile%_c = aggregate (con(nlcdep%tile% ge 0, 1),25) /* Calculate percent of each land use class at 1km resolution &do luc &list [unquote [translate [listunique nlcdep%tile%.vat -info value] ' ' ,]] &s tmpg [scratchname -dir] /* calculate (40m) cell count in this land use (luc) setcell 40 /* resample to 40m from 30m to make an even 1000m aggregate %tmpg% = aggregate(con(nlcdep%tile% eq %luc%,1),25) /* convert cell count to percent (at 1 km resolution) setcell 1000 nlcdep%tile%_%luc% = int ( ( 100.0 * %tmpg% / nlcdep%tile%_c ) + 0.5 ) kill %tmpg% all &end /* &do &list luc list &end /* &do for each tile /* merge together each 1k grid /* first set a mask for performance and snapping reset setcell 1000 /*Set window to nlcdep3_c for LL and nlcdep2_c for UR? Use extent of 1km NLCDE. setwindow -2380000 260000 2265000 3200000 /*-2360000 295000 2273000 3182000 &if ^ [exists xnlcdep_c -grid] &then ~ xnlcdep_c = mosaic(nlcdep1_c,nlcdep2_c,nlcdep3_c,nlcdep4_c) setwindow xnlcdep_c xnlcdep_c setmask xnlcdep_c /* merge 1km pct grids together - Exclude 12, 72 and 99 because not all 4 tiles are present for those &do luc &list 11 21 22 23 25 26 27 28 31 32 33 41 42 43 51 61 62 71 81 82 83 84 85 91 92 nlcdep_%luc% = mosaic(nlcdep1_%luc%,nlcdep2_%luc%,nlcdep3_%luc%,nlcdep4_%luc%) &end /*Put the oddball ones together setmask xnlcdep_c /*usg ?? &if [exists nlcdep_12 -grid] &then kill nlcdep_12 all /* class 12 has no grids for 2 and 4 (no snow and ice down south>) /* so we need to handle it separately x12 = mosaic(nlcdep1_12,nlcdep3_12) nlcdep_12 = setnull(isnull(xnlcdep_c),con(isnull(x12),0,x12)) kill x12 all &if [exists nlcdep_72 -grid] &then kill nlcdep_72 all /* class 72 has no grids for 2 and 4 (no snow and ice down south>) /* so we need to handle it separately x72 = mosaic(nlcdep1_72,nlcdep3_72) nlcdep_72 = setnull(isnull(xnlcdep_c),con(isnull(x72),0,x72)) kill x72 all &if [exists nlcdep_99 -grid] &then kill nlcdep_99 all /* class 99 has no grid for 1 /* so we need to handle it separately x99 = mosaic(nlcdep2_99,nlcdep3_99,nlcdep4_99) nlcdep_99 = setnull(isnull(xnlcdep_c),con(isnull(x99),0,x99)) kill x99 all quit &ret /*********************************************************** &routine cr81kmmajg &type Create majority grids grid setcell 1000 setwindow xnlcdep_c xnlcdep_c setmask xnlcdep_c /* Create majority land use class grid &if [exists nlcdep_clsg -grid] &then kill nlcdep_clsg all nlcdep_m = max([listfile nlcdep_* -grid]) nlcdep_clsg = con( ~ nlcdep_11 eq nlcdep_m, 11, ~ nlcdep_12 eq nlcdep_m, 12, ~ nlcdep_21 eq nlcdep_m, 21, ~ nlcdep_22 eq nlcdep_m, 22, ~ nlcdep_23 eq nlcdep_m, 23, ~ nlcdep_25 eq nlcdep_m, 25, ~ nlcdep_26 eq nlcdep_m, 26, ~ nlcdep_27 eq nlcdep_m, 27, ~ nlcdep_28 eq nlcdep_m, 28, ~ nlcdep_31 eq nlcdep_m, 31, ~ nlcdep_32 eq nlcdep_m, 32, ~ nlcdep_33 eq nlcdep_m, 33, ~ nlcdep_41 eq nlcdep_m, 41, ~ nlcdep_42 eq nlcdep_m, 42, ~ nlcdep_43 eq nlcdep_m, 43, ~ nlcdep_51 eq nlcdep_m, 51, ~ nlcdep_61 eq nlcdep_m, 61, ~ nlcdep_62 eq nlcdep_m, 62, ~ nlcdep_71 eq nlcdep_m, 71, ~ nlcdep_72 eq nlcdep_m, 72, ~ nlcdep_81 eq nlcdep_m, 81, ~ nlcdep_82 eq nlcdep_m, 82, ~ nlcdep_83 eq nlcdep_m, 83, ~ nlcdep_84 eq nlcdep_m, 84, ~ nlcdep_85 eq nlcdep_m, 85, ~ nlcdep_91 eq nlcdep_m, 91, ~ nlcdep_92 eq nlcdep_m, 92, ~ nlcdep_99 eq nlcdep_m, 99) /*kill nlcdep_m all /*kill xnlcdep_c all /*kill nlcdep(!1 2 3 4!)_c all &ret
NLCDEP CLASSES (modified from NLCD classification)
* = new NLCDEP class, added by overlaying NLCDE and 2000 population data.
+ = NLCDE class, added by overlaying NLCD and GIRAS data.
All other classes are standard NLCD categories.
VALUE LAND COVER
----- -------------------------------------------------------
11 Open Water
12 Perennial Ice/Snow
21 Low Intensity Residential
22 High Intensity Residential
23 Commercial/Industrial/Transportation
+ 25 GIRAS residential (but not NLCD water, urban, or forest)
+ 26 GIRAS residential (and NLCD forest)
* 27 Urban 2000 without trees
* 28 Urban 2000 with trees
31 Bare Rock/Sand/Clay
32 Quarries/Strip Mines/Gravel Pits
33 Transitional
41 Deciduous Forest
42 Evergreen Forest
43 Mixed Forest
51 Shrubland
61 Orchards/Vineyards/Other
+ 62 Orchards/Vinyards (burned from GIRAS)
71 Grasslands/Herbaceous
+ 72 Tundra (burned from GIRAS)
81 Pasture/Hay
82 Row Crops
83 Small Grains
84 Fallow
85 Urban/Recreational Grasses
91 Woody Wetlands
92 Emergent Herbaceous Wetlands
Fields in the percent grids (example shown is from "nlcde_11" grid) are:
nlcde_11.VAT:
COLUMN ITEM NAME WIDTH OUTPUT TYPE N.DEC
1 VALUE 4 10 B -
5 COUNT 4 10 B -
VALUE is the percentage of land use, Open water (code "11") in
the 1 square kilometer cell, rounded to integer values.
COUNT is the number of cells in the grid that have the corresponding VALUE.
Fields in the majority grid, "nlcde_clsg" are:
nlcde_CLSG.VAT
COLUMN ITEM NAME WIDTH OUTPUT TYPE N.DEC
1 VALUE 4 10 B -
5 COUNT 4 10 B -
VALUE is the land use code (e.g. 11 for "open water) that makes
up the highest percentage in the 1 square kilometer cell
COUNT is the number of cells in the grid that have the corresponding VALUE.
kjhitt@usgs.gov
kjhitt@usgs.gov