[OKGIS] FW: Conus NLCD2001 ftp Download

[Robert Stokes]

FYI
 
Robert Stokes
GIS Administrator
Cherokee Nation GeoData Center
P.O. Box 948
Tahlequah, OK 74465
Office: 918-453-5625
Cell: 918-207-2798
Fax: 918-458-7647
robert-stokes at cherokee.org
www.cherokee.org
 

GIS Workgroup et al.,

CONUS NLCD 2001 has been completed.  As you probably already know, the
most efficient way to get the data on your machine is to use the ftp
download at www.mrlc.gov.  NLCD 2001 data are delivered in "super"
mapping zones through the ftp download.  The embedded picture shows how
the "superzones" coincide with state boundaries.  Also attached is a
boundary coverage for the superzones.   The impervious surface and
canopy densities database elements can also be accessed at the ftp
download site.

Let me know if there are any questions.  Wendy may have sent this
announcement last week, so I apologize if this is redundant information.

---------
James (Jim) Wickham
Senior Research Biologist
US EPA, NERL (E243-05)
RTP, NC 27711
919.541.3077 (ph)
919.541.4329 (fx)
wickham.james at epa.gov

[attachment "superzones.e00" deleted by David Parrish/R6/USEPA/US]

NLCD 1992 - NLCD 2001 Change

Mapping methods, legends, and geometries between the two land-cover
datasets are sufficiently different that change detection should be
undertaken with caution.  Pixel-to-pixel comparison is not recommended.
Change should be examined on a proportional basis (e.g., using counties,
watersheds) at Anderson Level I (Water, urban, forest, agriculture,
shrub/grass).  The impervious surface mapping for NLCD 2001 yielded
significant differences in the resolution of roads between the two
datasets.  A road dataset should be "burned" into both land-cover
datasets prior to looking at proportional differences, so that change is
not confounded by the improved resolution of roads in the 2001
land-cover data.

Primary FY07 Activities

   Development of a "retrofitted" NLCD 1992 land-cover dataset for more
   direct change comparison with NLCD 2001.  Comparisons will still only
   be resolvable at Anderson Level I and road data may be needed  to
   make the two datasets more comparable.
   Accuracy assessment of NLCD 2001.  At this time, it is planned that
   there will be 4 primary attributes that will be assessed for thematic
   accuracy: 1) impervious surface, 2) canopy density, 3) land cover,
   and 4) proportional land cover estimates.  Item 3 represents the
   traditional per pixel (or minimum mapping unit) assessments that are
   reported in contingency tables.  The objective of Item 4 is to
   document the accuracy of percentage of forest, urban, and
   agriculture, which are often used in models and other assessments.
   Reconciling the necessarily different designs for items 3 and 4 will
   be a significant activity for FY07.
   Acquisition of Landsat 5 imagery for the target year 2006, and
   planning for a 5-year repeat cycle for NLCD.  Currently, it is
   envisioned that NLCD 2006 will be undertaken in a very similar
   fashion to the NOAA C-CAP program - spectral algorithms will be used
   to identify change, and only change will be mapped for land-cover in
   2006, leaving NLCD 2001 as a baseline.