tbohn / nlcd_inegi

NLCD_INEGI: Harmonized US-Mexico Land Cover Classifications from NLCD and INEGI, 1992/2001/2011

NLCD_INEGI: Harmonized US-Mexico Land Cover Classifications from NLCD and INEGI, 1992/2001/2011

This project contains scripts used to created harmonized US-Mexico land cover classifications for years 1992, 2001, and 2011.

Processing stages:

1. Harmonize the original INEGI Uso del Suelo y Vegetacion classifications (for the Mexican portion of the domain) with the NLCD classifications, and rasterize them.

• Inputs: the INEGI Uso del Suelo y Vegetacion classifications, SERIE I - V, which represent years 1985, 1993, 2002, 2007, and 2011.
• Outputs: the outputs are archived on Zenodo.

2. Merge the harmonized INEGI rasters with the NLCD products for complete coverage over the US and Mexico.

• Inputs: For Mexico, use the output of stage 1 (either run stage 1 or download outputs from Zenodo.). For the continental US, the NLCD land cover products for years 1992, 2001, and 2011, plus the NLCD 1992/2001 land cover change retrofit product, available from MRLC.
• Outputs: the outputs are archived on Zenodo

Contents:

1. tools/

• INEGI/ - arcPy scripts for stage 1, i.e., harmonizing the INEGI Uso del Suelo y Vegetacion classifications with the NLCD 2011 classification categories, reprojecting to geographic, and rasterizing.
• NLCD_INEGI/ - Perl scripts for stage 2, i.e., constructing an NLCD 1992 product consistent with the classifications of NLCD 2001 and 2011, merging the NLCD maps with the INEGI maps, and creating the "stable_2001-2011" dataset.
• hybrid/ - Perl script for creating optional hybrid land cover map, in which a land use map focused on urban land use (potentially with future projections) is draped on top of the NLCD_INEGI map.

2. docs/

• Processing_of_INEGI_USOSV_dataset.docx - Description of procedure for harmonizing the INEGI Uso del Suelo y Vegetacion classifications with the NLCD 2011 classification categories, reprojecting to geographic, and rasterizing.
• Processing_NLCD_INEGI.md - Description of procedure for constructing an NLCD 1992 product consistent with the classifications of NLCD 2001 and 2011, and for merging the NLCD maps with the INEGI maps.
• Processing_hybrid.md - Description of procedure for constructing a hybrid of urban-focused land use map draped on top of the NLCD_INEGI map.

3. examples/

• Codes.ICLUS.csv - example land use table for use in creating hybrid land cover map.

nlcd_inegi's Issues

NLCD resolution

The "process_urban_lu_map.pl" thinks that the spatial resolution of NLCD map is 0.0003508840696170 degree. I'm not sure why this values was selected. But, this is not always the case. For example, the resolution I used is 0.000341219290263889 degree and this caused problem in the overlaying step. It'd be more reasonable to read the spatial resolution from the header of the nlcd map or from user input.

issues with Process_hybrid tools and the documentation

1. "mix" comment in the example $LU_CODES_TABLE file
   The "process_urban_lu_map.pl" script uses "mix" to identify those class codes that need to be replaced with NLCD_INEGI class codes. However, the example $LU_CODES_TABLE file uses "mixed" , which might be misleading to users.

2. wrong column number
   The "process_urban_lu_map.pl" reads column 3 of the $LU_CODES_TABLE file to decide if the class needs to be replaced or not. However, the example $LU_CODES_TABLE file puts the "nulled"/"mix" comment in the 4th column. Thus, the $fields[2] should be changed to $fields[3] in line 37, 40 and 43.

3. file name rule of the "$LAND_USE_FILE"
   The first argument read by the "process_urban_lu_map.pl" script ("$LAND_USE_FILE" ) needs to be end with ".resolution_deg.asc". It'd better to specify this in the documentation file.