BOREAS AFM-12 1-km AVHRR Seasonal Land Cover Classification Summary The BOREAS AFM-12 team’s efforts focused on regional scale SVAT modeling to improve parameterization of the heterogeneous BOREAS landscape for use in larger scale GCM’s. This regional land cover data set was developed as part of a multitemporal 1-km AVHRR land cover analysis approach that was used as the basis for regional land cover mapping, fire disturbance-regeneration, and multiresolution land cover scaling studies in the boreal forest ecosystem of central Canada (Steyaert et al., 1997). This land cover classification was derived by using regional field observations from ground and low-level aircraft transits to analyze spectral-temporal clusters that were derived from an unsupervised cluster analysis of monthly NDVI image composites (April-September 1992). This regional data set was developed for use by BOREAS investigators, especially those involved in simulation modeling, remote sensing algorithm development, and aircraft flux studies. Based on regional field data verification, this multitemporal 1-km AVHRR land cover mapping approach was effective in characterizing the biome-level land cover structure, embedded spatially heterogeneous landscape patterns, and other types of key land cover information of interest to BOREAS modelers. The land cover mosaics in this classification include: (1) wet conifer mosaic (low, medium, and high tree stand density), (2) mixed coniferous-deciduous forest (80% coniferous, codominant, and 80% deciduous), (3) recent visible burn, vegetation regeneration, or rock outcrops-bare ground-sparsely vegetated slow regeneration burn (4 classes), (4) open water and grassland marshes, and (5) general agricultural land use/grasslands (3 classes). This land cover mapping approach did not detect small subpixel-scale landscape features such as fens, bogs, and small water bodies. Field observations and comparisons with Landsat TM suggest a minimum effective resolution of these land cover classes in the range of 3 to 4 km, in part, because of the daily to monthly compositing process. In general, potential accuracy limitations are mitigated by the use of conservative parameterization rules such as aggregation of predominant land cover classes within minimum horizontal grid cell sizes of 10 km. More detailed discussion is provided by Steyaert et al. (1997). The data are stored in binary image format files. Table of Contents * 1 Data Set Overview * 2 Investigator(s) * 3 Theory of Measurements * 4 Equipment * 5 Data Acquisition Methods * 6 Observations * 7 Data Description * 8 Data Organization * 9 Data Manipulations * 10 Errors * 11 Notes * 12 Application of the Data Set * 13 Future Modifications and Plans * 14 Software * 15 Data Access * 16 Output Products and Availability * 17 References * 18 Glossary of Terms * 19 List of Acronyms * 20 Document Information 1. Data Set Overview This regional land cover classification is based on the use of multitemporal 1- km Advanced Very High Resolution Radiometer (AVHRR) National Oceanic and Atmospheric Administration (NOAA 11) data that were analyzed in combination with selected Landsat Thematic Mapper (TM) and extensive field observations within a 619-km by 821-km subset of the 1,000-km by 1,000-km BOReal Ecosystem-Atmosphere Study (BOREAS) region (Steyaert et al., 1997). Following the approach developed by Loveland et al. (1991) for 1-km AVHRR land cover mapping in the conterminous United States, monthly Normalized Difference Vegetation Index (NDVI) image composites (April-September 1992) of this subset in the BOREAS region were used in an unsupervised image cluster analysis algorithm to develop an initial set of seasonal land cover classes. Extensive ground data with Global Positioning System (GPS) georeferencing, observations from low-level aerial flights over remote areas, and selected Landsat image composites for the study areas were analyzed to split, aggregate, and label the spectral-temporal clusters throughout the BOREAS region. Landsat TM image composites (bands 5, 4, and 3) were available for the 100-km by 100-km Northern Study Area (NSA) and Southern Study Area (SSA). This AVHRR land cover product was compared with Landsat TM land cover classifications for the BOREAS study areas (Steyaert et al., 1997). 1.1 Data Set Identification BOREAS AFM-12 1-km AVHRR Seasonal Land Cover Classification 1.2 Data Set Introduction This data set consists of a regional land cover analysis for a 619-km by 821-km subset of the BOREAS region. This experimental land cover data set was developed for test and evaluation as part of regional modeling and field experiments in BOREAS. 1.3 Objective/Purpose A major objective of this study was to develop a regional 1-km AVHRR land cover classification for use by BOREAS investigators. The development and intercomparison of advanced multiresolution land cover mapping techniques based on Landsat TM and AVHRR data was a secondary objective of this research. 1.4 Summary of Parameters The types of forest land cover classes required for input to BOREAS Terrestrial Ecology (TE) models were identified at a joint meeting of TE modelers and Remote Sensing Science (RSS) algorithm developers held in Columbia, MD, during June 1993. These general forest land cover classes, subsequently endorsed by other BOREAS modeling groups (Airborne Fluxes and Meteorology (AFM), Tower Fluxes (TF), and Trace Gas Biogeochemistry (TGB)), include Wet Conifer, Dry Conifer, Mixed Forest (Coniferous and Deciduous), Deciduous, Disturbed, Fen, Water, Regeneration (young, medium, and old age categories), and recent burn areas. Landsat TM 30-m data were used in the BOREAS TE-18 investigation to develop a land cover classification for this classification scheme in both the NSA and SSA in the BOREAS region (see F. Hall and D. Knapp, TE-18 documentation). The land cover classes defined as part of this regional 1-km AVHRR land cover classification are in general conformance with the types of land cover classes developed from Landsat TM classifications for the NSA and SSA. The differences between the classifications are, in part, associated with the different resolutions of the satellite sensors; that is, the 1-km AVHRR pixel size versus the 30 m Landsat TM pixel size. For example, this regional land cover mapping approach did not detect subpixel fens, bogs, and small lakes. There are also additional land cover classes that are in the BOREAS region, but not in the study areas. These 1-km AVHRR-derived land cover classes for BOREAS generally represent mosaics of various vegetation species as described in Section 7.3.2. The 1-km AVHRR land cover classes for BOREAS include: Class ID Class Name 1 Wet Conifer (Low Stand Density) 2 Wet Conifer (Medium Stand Density) 3 Wet Conifer (High Stand Density) 4 Upland Conifer/Fen 5 Rock Outcrops/Bare Ground/Sparse Vegetation/ Slow Regeneration Burn Areas 6 NA 7 Open Water 8 NA 9 Regeneration (North: Within Canadian Shield Zone) 10 NA 11 Recent Visible Burn 12 Rangeland/Pasture/Hay/Aspen Patches 13 Mixed Agriculture/Predominately Grains 14 Mixed Agriculture/Predominately Pasture/Hay 15 Grassland Marshes 16 Mixed Forest (80% Coniferous) 17 Mixed Forest (50% Coniferous) 18 Mixed Forest (80% Deciduous) 19 Regeneration (South: generally south of Shield Zone) 20 Unknown 1.5 Discussion This experimental land cover data base for BOREAS was developed from multitemporal 1-km AVHRR data that were supplemented by field observations and selected Landsat TM image composites for the two study areas in BOREAS (Steyaert et al., 1997). The preprocessing of the daily 1-km AVHRR data and the monthly NDVI unsupervised cluster analysis followed procedures outlined by Eidenshink (1992a, b), Loveland et al. (1991), and Brown et al.(1993). Field observations, collected during the pre-BOREAS operations in 1993 and BOREAS Intensive Field Campaigns (IFCs) of 1994, were the primary source of information to analyze, combine, and interpret the clusters according to land cover class. These field data were the primary source of information for the analysis of regional forest fire disturbance-regenerating vegetation patterns. The 1-km AVHRR land cover classification was compared with both high- and coarse-resolution land cover data bases generated from 30-m Landsat TM and AVHRR data, respectively. 1.6 Related Data Sets BOREAS TE-18 Landsat TM Physical Classification Image of the NSA BOREAS TE-18 Landsat TM Physical Classification Image of the SSA BOREAS TE-18 Landsat TM Maximum Likelihood Classification Image of the NSA BOREAS TE-18 Landsat TM Maximum Likelihood Classification Image of the SSA 2. Investigator(s) 2.1 Investigator(s) Name and Title Dr. Louis T. Steyaert Remote Sensing Scientist USGS EROS Data Center Dr. Forrest G. Hall NASA GSFC Dr. Thomas R. Loveland Remote Sensing Scientist USGS EROS Data Center 2.2 Title of Investigation Modeling Biosphere-Atmosphere Interactions at Various Scales in Support of BOREAS (AFM-12, PI: Dr. R.A. Pielke, Sr.) 2.3 Contact Information Contact 1 ----------------- Dr. Lou Steyaert USGS EROS Data Center NASA GSFC Greenbelt, MD (301) 286-2111 (301) 286-0239 steyaert@ltpmail.gsfc.nasa.gov Contact 2 ----------------- David Knapp Raytheon STX Corporation NASA GSFC Greenbelt, MD (301) 286-1424 (301) 286-0239 David.Knapp@gsfc.nasa.gov 3. Theory of Measurements The AVHRR is a four- or five-channel scanning radiometer capable of providing global daytime and nighttime information about ice, snow, vegetation, clouds, and the sea surface (Newcomer, 1992; Los et al., 1995). These data are obtained on a daily basis primarily for use in weather analysis and forecasting; however, a variety of other applications are possible. The AVHRR data are from instruments onboard the NOAA polar orbiting platforms. The radiometers onboard more recent satellites such as NOAA-11 measure emitted and reflected radiation in two visible, one middle infrared, and two thermal channels. The spectral regions, bandwidths, and primary use of each channel are given in the following table: [micro- Channel Wavelength meters] Primary Use ------- ------------------- --------------------------------- 1* 0.58 - 0.68 Daytime Cloud and Surface Mapping 2 0.725- 1.10 Surface Water Delineation Vegetation Cover 3 3.55 - 3.93 Sea Surface Temperature (SST) Nighttime Cloud Mapping 4** 10.5 - 11.5 Surface Temperature, Day/Night Cloud Mapping 5*** 11.5 - 12.5 Surface Temperature * Channel 1 wavelength for Television and Infrared Observation Satellite (TIROS)-N flight model was 0.55 - 0.90 micrometers. **For NOAA-7 and -9 Channel 4 was 10.3 - 11.3 micrometers. ***For TIROS-N, NOAA-6, -8, and -10, Channel 5 is a duplicate of Channel 4. The wavelength ranges at 50% Relative Spectral Response (in micrometers) of the bands for each platform are: Band NOAA-9 NOAA-10 NOAA-11 ------ --------------- --------------- --------------- 1 0.570 - 0.699 0.571 - 0.684 0.572 - 0.698 2 0.714 - 0.983 0.724 - 0.984 0.716 - 0.985 3 3.525 - 3.931 3.554 - 3.950 3.536 - 3.935 4 10.334 - 11.252 10.601 - 11.445 10.338 - 11.287 5 11.395 - 12.342 10.601 - 11.445 11.408 - 12.386 The AVHRR is capable of operating in both real-time or recorded modes. Direct readout data can be transmitted to ground stations of the automatic picture transmission (APT) class at low resolution (4 x 4 km) and to ground stations of the high-resolution picture transmission (HRPT) class at high resolution (1 x 1 km), such as the HRPT receiving and processing station at the United States Geological Survey (USGS) Earth Resources Observation System (EROS) Data Center (EDC), Sioux Falls, SD. Data recorded onboard are available for processing after downlinking to groundstations such as at the USGS-EDC Data Center or at the Naval Research Laboratory (NRL) Satellite Data Receiving and Processing Facility. These recorded data include global area coverage (GAC) data, with a resolution of 4x4 km, and local area coverage (LAC) data recorded from selected portions of each orbit with a 1 x 1 km resolution. 4. Equipment 4.1 Sensor/Instrument Description The AVHRR onboard NOAA-11 is a cross-track scanning system featuring two visible, one middle infrared, and two thermal channels. 4.1.1 Collection Environment The daily NOAA-11 AVHRR data were collected by the USGS EDC during the period 1- April - 30-September-1992. Field study trips were made to various sites in the BOREAS region during July 19, 1993, July 19, 1994, August 19, 1994, and July 19, 1996. 4.1.2 Source/Platform AVHRR data used for this BOREAS data set were collected onboard the NOAA-11 polar orbiting platform. The NOAA-11 is an afternoon pass satellite with northbound Equatorial crossing directly after launch of 1340 LST. However, during the time of operation of the satellite, the Equatorial crossing time gradually shifted to a later time in the afternoon. 4.1.3 Source/Platform Mission Objectives The AVHRR is designed for multispectral analysis of meteorological, oceanographic, and hydrologic parameters. The objective of the instrument is to provide radiance data for investigation of clouds, land-water boundaries, snow and ice extent, ice or snow melt inception, day and night cloud distribution, temperatures of radiating surfaces, and SST. It is an integral member of the payload on the advanced TIROS-N spacecraft and its successors in the NOAA series, and as such contributes data required to meet a number of operational and research-oriented meteorological objectives. Although not its primary purpose, the AVHRR was found to be suitable for vegetation monitoring studies, in part because of its high temporal resolution and global coverage. 4.1.4 Key Variables Emitted radiation. Reflected radiation (used to calculate the NDVI). 4.1.5 Principles of Operation The AVHRR is a four- or five-channel scanning radiometer that detects emitted and reflected radiation from Earth in the visible, near-infrared, and far- infrared regions of the spectrum. A fifth channel has been added to the follow- on instrument designated AVHRR/2 and flown on NOAA-7, NOAA-9, and NOAA-11 to improve the correction for atmospheric vapor. Scanning is provided by an elliptical beryllium mirror rotating at 360 rpm about an axis parallel to Earth’s axis. A two-stage radiant cooler is used to maintain a constant temperature for the infrared detectors of 95 K. The operating temperature is selectable at either 105 or 110 K. The telescope is an 8-inch afocal, all- reflective Cassegrain system. Polarization is less than 10 percent. Instrument operation is controlled by 26 commands and monitored by 20 analog housekeeping parameters. 4.1.6 Sensor/Instrument Measurement Geometry The AVHRR is a cross-track scanning system. The instantaneous field-of-view (IFOV) of each sensor is approximately 1.4 milliradians giving a resolution of 1.1 km at the satellite subpoint. There is about a 36 percent overlap between IFOVs (1.362 samples per IFOV). The scanning rate of the AVHRR is six scans per second, and each scan spans an angle of +/- 55.4 degrees from the nadir. 4.1.7 Manufacturer of Sensor/Instrument ITT Aerospace/Communications Division P.O. Box 3700 Fort Wayne, IN 46801-3700 4.2 Calibration NOAA provides calibration parameters on tape that relate the data in the visible and near-infrared channels to a preflight standard (preflight calibration). However, during the time of operation of the satellite, the sensitivity of the red and near-infrared has gradually decreased. This decrease is not accounted for by the preflight calibration, and no in-flight visible channel calibration is performed. The calibration coefficients for AVHRR thermal channels 3, 4, and 5 are derived onboard the satellite using a view of a stable blackbody and deep space as a reference. The radiance values for all channels are stored with 10- bit precision. The procedures used by the USGS EDC to radiometrically calibrate the visible and near-infrared channels as well as other methods used by EDC to process these data are described by Eidenshink (1992a, 1992b). Specific details are provided in Section 4.2.2. 4.2.1 Specifications IFOV 1.4 mRad RESOLUTION 1.1 km ALTITUDE 833 km SCAN RATE 360 scans/min 1.362 samples per IFOV SCAN RANGE -55.4 to 55.4 degrees SAMPLES/SCAN 2048 samples per channel per Earth scan 4.2.1.1 Tolerance The AVHRR infrared channels were designed for a Noise Equivalent Differential Temperature (NEdt) of 0.12 Kelvin (at 300 Kelvin, and a signal-to-noise ratio of 3:1 at 0.5 percent albedo. 4.2.2 Frequency of Calibration NOAA provides calibration parameters on tape that relate the data to a preflight standard (preflight calibration). These parameters generally do not change during the time of operation of a satellite (with the exception of NOAA-11). The preflight calibration does not take degradation of the sensors into account. Degradation of AVHRR sensors after launch is well documented (e.g., Rao, 1987; Price, 1987; Holben et al., 1990). These studies have used a variety of approaches such as ground-based measurements from stable sites (e.g., homogeneous desert targets) to monitor the degradation of the sensors. Corrections to these 1992 data for sensor degradation were made by using coefficients developed from a study by Teillet and Holben (1992, unpublished report) and Teillet and Holben (1994). Their calculation takes into account the desert calibration approach (Holben and et al., 1990) to develop a set of time- dependent calibration coefficients for the AVHRR sensor on NOAA-11. Use of calibration coefficients involves extrapolation of the most recent calibration results for processing data on a near-real-time basis. Therefore, the time-dependent coefficients are based on a piecewise linear fit of the desert results. A piecewise linear fit is recommended for operational use because, unlike polynomial fits, it will not change retroactively when new data are added to the end of the time series (Eidenshink, 1992). 4.2.3 Other Calibration Information See Eidenshink (1992a, 1992b). 5. Data Acquisition Methods The primary data source for this 1-km AVHRR land cover classification was monthly NDVI image composites derived from daily NOAA-11 AVHRR polar orbiting satellite data. Daily 1-km AVHRR data were received and processed by the USGS EDC during the period April through September 1992. Monthly NDVI composites were processed for a 619-km by 821-km subset (approximately 52-57 N and 96-108 W) of the 1,000-km by 1,000-km BOREAS region. The methods for processing the daily NOAA-11 AVHRR data into monthly NDVI image composites is described by Eidenshink (1992a, 1992b). These processing steps include radiometric calibration, atmospheric correction, computation of the NDVI, geometric registration, and image compositing. Other than the maximum NDVI compositing, no cloud screening algorithm was used. The AVHRR data were not atmospherically corrected for water vapor and aerosols. These processing procedures are very analogous to the procedures established for processing the 1-km AVHRR Pathfinder data as described by Eidenshink and Faundeen (1994). Landsat imagery and field observations were essential data inputs to the development of this land cover classification. Hardcopy Landsat TM image composites (bands 5, 4, and 3) for the NSA and the SSA were acquired from TE-18 to help understand and identify 1-km AVHRR spectral-temporal clusters in the study areas and as a guide to help interpret field observations in the study areas. Extensive field observations of vegetation type and composition along with associated GPS georeferencing data were obtained throughout many portions of the BOREAS region. A four-wheel drive vehicle was used to collect more than 350 sets of ground observations with GPS positional fixes, mainly along the road networks within the boreal forested areas of the BOREAS region during field visits (July 1993; July and September 1994). In addition, several low-level aircraft reconnaissance flights within Saskatchewan and Manitoba, including the BOREAS transect from Prince Albert National Park (PANP) to Thompson, were used to photograph vegetation patterns (some GPS) and verify preliminary classes. These aircraft flights by both USGS and NASA investigators were especially useful in analysis of land cover conditions in remote regions. A follow-up field visit including three separate low-level aircraft flights in Manitoba was made during July 1996. 6. Observations 6.1 Data Notes None. 6.2 Field Notes Extensive field observations of land cover type composition were made during 1993 and 1994 with a follow-up visit in 1996. 7. Data Description 7.1 Spatial Characteristics 7.1.1 Spatial Coverage The regional 1-km AVHRR land cover data are located in a 672 row by 862 column raster image. This image contains the actual land cover classes (pixel values 1-20) for the 619-km by 821-km subset of the BOREAS region, plus a set of zero- value pixels that form the boundary of the raster image. The subsetted land cover classification has a domain of approximately 52-57 deg. N and 96-108 deg. W, which includes the BOREAS SW-NE transect from southwest of Saskatoon, Saskatchewan, to northeast of Gillam, Manitoba. The corners of the data set are as follows. These coordinates are in the BOREAS Grid projection. Corner X Y ------------------------------------ Upper Left 174.0707 785.4531 Upper Right 1036.0707 785.4531 Lower Left 174.0707 113.4531 Lower Right 1036.0707 113.4531 7.1.2 Spatial Coverage Map None. 7.1.3 Spatial Resolution The IFOV of each sensor is approximately 1.4 milliradians, leading to a resolution of about 1.1 km by 1.1 km at nadir for a nominal altitude of 833 km. The AVHRR and land cover data were gridded to a cell size of 1.0 km from the original nominal resolution of 1.1 km. However, as discussed by Steyaert et al. (1997) the effective resolution of these 1-km AVHRR land cover classes is more realistically in the 3-4 km range, in part, because of the maximum NDVI compositing and multitemporal analysis over the April-September time period. 7.1.4 Projection The area mapped is projected in the Albers Equal Area Conic (AEAC) projection. The projection has the following parameters: Datum: None Ellipsoid: Sphere Origin: 111.000 ?W 51.000 ?N Standard Parallels: 52? 30' 00" N 58? 30' 00" N Units of Measure: kilometers Note: Each pixel is 1,000 m by 1,000 m. It is important to emphasize that this image is projected using a Sphere as the Earth model and not the WGS84 ellipsoid used for most other BOREAS data sets. The other projection parameters listed above are the same as many other BOREAS georeferenced data sets. This difference in ellipsoid models can result in spatial misregistration of approximately 2 to 4 pixels. This difference should be considered when comparing this classification to other georeferenced imagery. 7.1.5 Grid Description None. 7.2 Temporal Characteristics 7.2.1 Temporal Coverage Monthly NDVI image composites for the period April-September 1992 were used to develop this 1992 1-km AVHRR/land cover data set. 7.2.2 Temporal Coverage Map None. 7.2.3 Temporal Resolution Monthly NDVI image composites for the period April-September, 1992 were used to develop this 1992 1-km AVHRR/land cover data set. 7.3 Data Characteristics 7.3.1 Parameter/Variable Land Cover Type. 7.3.2 Variable Description/Definition The 1-km AVHRR land cover classes listed in Section 1.4 can be grouped into broad vegetation categories consisting of: (1) wet conifer mosaic; (2) mixed coniferous-deciduous forest; (3) recent burn, regeneration, or rock outcrops- bare ground-sparsely vegetated slow regeneration burn; (4) open water and grassland marshes; and (5) general agricultural land use. This grouping facilitates the understanding and descriptive characterization of these 1-km AVHRR land cover classes, especially in terms of the vegetation associations within each class. Qualitative descriptions of these land cover classes are essential for interpreting class attributes and estimating biophysical parameters for land surface parameterizations. The land cover class descriptions within these vegetation categories are given in the following subsections: Wet Conifer Mosaic The wet conifer mosaic is the dominant conifer class within this classification. This wet conifer mosaic consists of black spruce (Picea mariana) and various embedded subpixel fens and bogs, scattered tamarack (Larix laricina), mixed water-vegetation pixels, small pockets of dry jack pine (Pinus banksiana) on sandy hilltops, and scattered deciduous trees. This mosaic is characterized by the very consistent vegetation patterns in the "lowlying" areas (black spruce, fens, bogs) as opposed to more upland terrain (more productive black spruce in combination with jack pine on sandy soils and scattered deciduous trees) environments throughout the entire BOREAS region. This classification does not resolve in all cases these "lowland" versus "upland" components of the wet conifer mosaic. The subpixel fens, bogs, and small water bodies are also not resolved in this classification. Based on extensive field data, the 1-km AVHRR spectral-temporal clusters do permit the characterization of the wet conifer mosaic into "low,” "medium,” and "high" tree density levels (Classes 1-3, respectively). There is also a small upland conifer/fen class (Class 4) that is characterized by isolated patches of mature jack pine or black spruce/fen mosaics. This class is in part due to the lack of spectral separation between dense black spruce and jack pine classes with AVHRR. To the east of Lake Winnipeg, this mixed conifer mosaic consists of black spruce (with some jack pine) on small, "upland" hummocks that are embedded in large tamarack fens. Mixed Coniferous-Deciduous Forest Mosaic There are three AVHRR mixed forest classes that, based on field observations, are estimated to consist of 80 percent conifer-20 percent deciduous (Class 16), codominant mixed forest (Class 17), and 80 percent deciduous-20 percent conifer (Class 18). These mixed forest classes are generally distributed along a southwest-northeast gradient ranging from deciduous dominant in the south to coniferous dominant in the north. The effects of forest succession are evident in this mixed class, especially in stands with mature deciduous trees and successional spruce under the deciduous canopy. In the northern extremes, this AVHRR mixed forest (Class 16) is predominantly upland black spruce with scattered jack pine on sandy soils and approximately 20 percent aspen trees (populus tremuloides) with scattered birch (betula papyrifera) and balsam poplar (populus balsamifera) trees. These trees are typically on rocky hills throughout the central and northern portions of the BOREAS region. The mixed forest class in the central region (Class 17) consists of codominant coniferous and deciduous trees that are quite well developed. The conifers are dominated by tall jack pine, black spruce, and some white spruce (Picea glauca), while the deciduous trees consist of mature aspen and birch. The mixed forest in the southern boreal ecosystem (Class 18) is dominated by well-developed aspen trees that grow either in pure stands or in mixed forest stands with birch, balsam poplar, and some conifers. The deciduous trees account for at least 80 percent of the mixed forest composition. In some cases, the aspen is near maturity with well developed conifers growing under the canopy. Recent Burn, Regeneration, or Rock Outcrops-Bare Ground-Sparsely Vegetated-Burn This grouping includes individual land cover classes for recent visible burns (Class 11), fire disturbance-regenerating vegetation in the north (Class 9) and south (Class 19), and a rock outcrops-bare ground-sparsely vegetated class (Class 5) that is frequently associated with slow regeneration burned areas of various ages. The recent visible burns (Class 11) represent areas that were burned within the past 5 or 6 years, relative to this 1992 data analysis. This burn class is distinguishable by its charred background of partially burned trees and moss in black spruce areas or other intensely burned areas where little or no vegetation survived. These recent burn areas are much more frequent and widespread within the Canadian Shield Zone. The regenerating vegetation patches in the north (Class 9) are located within the Canadian Shield Zone and are typically associated with old burns of various ages. This mixed vegetation class consists of jack pine, aspen, and young black spruce trees. The stand density and tree sizes depend on the age of the burn and the soil conditions. The jack pine and aspen trees are taller than young black spruce. The regenerating vegetation patches in the south (Class 19) that are located to the south of the Canadian Shield Zone are the result of old burns or previously logged or cleared areas, especially along the southern ecotone between the boreal forest and grasslands/agriculture land cover types. The class is dominated by a mixture of regenerating young aspen trees and various herbaceous bushes and grasses with scattered jack pine. The rock outcrops-bare ground- sparsely vegetated areas (Class 5) are frequently associated with slow regeneration burn areas of various ages, especially within the north. Based on field data analysis, this mixed class has examples of recent and older burns. The estimated vegetation cover for Class 5 is less than 30 percent. Open Water and Grassland Marshes The open water Class 7 includes water bodies such as small lakes and streams. For AVHRR, this class also includes water/vegetation mixed pixels. In the north, there were some cases of definite spectral confusion between dark water bodies and recent, dark burns. The grassland marshes (Class 15) are mainly located in western Manitoba, near The Pas. General Agricultural Land Use There are three general agricultural land use classes. A mixed class (rangeland, pasture, hay, and aspen patches) consists of aspen patches (typically around "pothole" water bodies) embedded in grasslands (rangeland, pasture, and hay fields) in the southwest portion of the BOREAS region (Class 12). The aspen trees are estimated to represent 30 percent of the land cover. A mixed agricultural, predominantly grains class represents the major agricultural grain-producing area in the BOREAS region (Class 13). This class also includes fallowed fields. A mixed agricultural class consists predominantly of pasture and hay fields with some grain cropping (Class 14). 7.3.3 Unit of Measurement Pixel values represent different land cover types. 7.3.4 Data Source AVHRR imagery was received from EDC, Sioux Falls, SD. Pixel values range from 0 to 20, where the land cover classes have pixel values of 1-20 and the raster-filler boundary pixels have values of zero. 7.3.5 Data Range Pixel values range from 0 to 20, where the land cover classes have pixel values of 1-20 and the raster-filler boundary pixels have values of zero. 7.4 Sample Data Record Not applicable to image data. 8. Data Organization 8.1 Data Granularity The smallest amount of data that can be ordered from this data set is the entire data set. 8.2 Data Format(s) 8.2.1 Uncompressed Data Files This BOREAS AFM-12 regional 1-km AVHRR/land cover classification product contains two files as follows: File 1: (80-byte American Standard Code for Information Interchange (ASCII) text records) Header file on tape File 2: (672 records of 862 bytes each) (1 byte per pixel) Classified image with values from 1 to 20 with zero values(0) used as fillers for the boundaries of the image. 8.2.2 Compressed CD-ROM Files On the BOREAS CD-ROMs, the ASCII header file for this image is stored as ASCII text; however, file 2 has been compressed with the Gzip compression program (file name *.gz). These data have been compressed using gzip version 1.2.4 and the high compression (-9) option (Copyright (C) 1992-1993 Jean-loup Gailly). Gzip (GNU zip) uses the Lempel-Ziv algorithm (Welch, 1994) used in the zip and PKZIP programs. The compressed files may be uncompressed using gzip (-d option) or gunzip. Gzip is available from many websites (for example, ftp site prep.ai.mit.edu/pub/gnu/gzip-*.*) for a variety of operating systems in both executable and source code form. Versions of the decompression software for various systems are included on the CD-ROMs. 9. Data Manipulations 9.1 Formulae None. 9.1.1 Derivation Techniques and Algorithms None. 9.2 Data Processing Sequence 9.2.1 Processing Steps The primary data source for this 1-km AVHRR land cover classification was monthly NDVI image composites derived from daily NOAA-11 AVHRR polar orbiting satellite data. Daily 1-km AVHRR data were received and processed by the USGS EDC during the period April through September 1992. Monthly NDVI composites were processed for a 619-km by 821-km subset of the BOREAS region (approximately 52- 57 N and 96-108 W). The methods for processing the daily NOAA-11 AVHRR into monthly NDVI image composites is described by Eidenshink (1992a, 1992b). These processing steps include radiometric calibration, atmospheric correction, computation of the NDVI, geometric registration, and image compositing. Other than the maximum NDVI compositing, no cloud screening algorithm was used. The AVHRR data were not atmospherically corrected for water vapor and aerosols. These processing procedures are very analogous to the procedures established for processing the 1-km AVHRR Pathfinder data as described by Eidenshink and Faundeen (1994). This regional land cover classification was then based on the use of these multitemporal 1-km AVHRR (NOAA-11) data that were analyzed in combination with selected Landsat TM and extensive field observations within a 619 km by 821 km subset of the BOREAS region (Steyaert et al., 1997). Following the approach developed by Loveland et al. (1991) for 1-km AVHRR land cover mapping in the conterminous United States, monthly NDVI image composites (April-September 1992) of this subset in the BOREAS region were used in an unsupervised image cluster analysis algorithm to develop an initial set of seasonal land cover classes. Extensive ground data with GPS georeferencing, observations from low-level aerial flights over remote areas, and selected Landsat image composites for the study areas were analyzed to split, aggregate, and label the spectral-temporal clusters throughout the BOREAS region. Landsat TM image composites (bands 5, 4, and 3) were available for the 100-km by 100-km NSA and SSA. This AVHRR land cover product was compared with Landsat TM land cover classifications for the BOREAS study areas (Steyaert et al., 1997). 9.2.2 Processing Changes None. 9.3 Calculations None. 9.3.1 Special Corrections/Adjustments None. 9.3.2 Calculated Variables 9.4 Graphs and Plots None. 10. Errors 10.1 Sources of Error The sources of error in this classification could be the result of a number of factors. Error could be the result of spectral mixing of various features that fall within a 1-km pixel. The spectral signature of one feature could also be similar to that of another feature, resulting in confusion. The similarity in spectral signatures could be the result of similar background components and variations in tree density. The locational accuracy of this image may be off by as much as 3 or 4 pixels based on the compositing of the multitemporal data. 10.2 Quality Assessment 10.2.1 Data Validation by Source The imagery was spot checked at various locations, and the image class was compared to Landsat TM-derived land cover classifications developed for the NSA and SSA under TE-18. 10.2.2 Confidence Level/Accuracy Judgment Although efforts have been made to make this classification as accurate as possible, there is bound to be some confusion between classes. The most noticeable problem is confusion between dense jack pine and dense black spruce. Spectrally, they look very similar. 10.2.3 Measurement Error for Parameters Not applicable. 10.2.4 Additional Quality Assessments None. 10.2.5 Data Verification by Data Center This image was viewed to make sure that it matched the product description. 11. Notes 11.1 Limitations of the Data This product is intended to be used to characterize the land cover over a large region at least at a 1 km pixel resolution or greater. It should not be used to determine land cover at a few specific pixels. 11.2 Known Problems with the Data The AVHRR data were not atmospherically corrected for water vapor and aerosols. This could have affected the NDVI values that were derived from the data, and thus this classification. 11.3 Usage Guidance Before uncompressing the Gzip files on CD-ROM, be sure that you have enough disk space to hold the uncompressed data files. Then use the appropriate decompression program provided on the CD-ROM for your specific system. 11.4 Other Relevant Information None. 12. Application of the Data Set This data set would be useful for regional modeling of the ecosystem. 13. Future Modifications and Plans None given. 14. Software 14.1 Software Description Not applicable. 14.2 Software Access Not applicable. 15. Data Access 15.1 Contact Information Primary contact: Ms. Beth Nelson BOREAS Information System NASA GSFC Greenbelt, MD (301) 286-4005 (301) 286-0239 Elizabeth.Nelson@gsfc.nasa.gov 15.2 Data Center Identification See Section 15.1 15.3 Procedures for Obtaining Data Users may place data requests by telephone, electronic mail, or fax. 15.4 Data Center Status/Plans The AFM-12 AVHRR land cover classification data are available from Earth Observing System Data and Information System (EOSDIS) Oak Ridge National Laboratory (ORNL), Distributed Active Archive Center (DAAC). The BOREAS contact at ORNL is: ORNL DAAC User Services Oak Ridge National Laboratory (865) 241-3952 ornldaac@ornl.gov ornl@eos.nasa.gov 16. Output Products and Availability 16.1 Tape Products These data can be made available on 8-mm, Digital Archive Tape (DAT), or 9-track tapes at 6250 or 1600 Bytes Per Inch (BPI). 16.2 Film Products None. 16.3 Other Products None. 17. References 17.1 Platform/Sensor/Instrument/Data Processing Documentation Kidwell, K.B. 1991. NOAA Polar Orbiter Data Users' Guide. National Oceanic and Atmospheric Administration, World Weather Building, Room 100, Washington, D.C. PACE Image Analysis Kernal Version 5.2. 1993. PCI Inc. Richmond Hill, Ontario. Richards, J.A. 1986. Remote Sensing Digital Image Analysis: An Introduction. Springer Verlag. Welch, T.A. 1984, A Technique for High Performance Data Compression, IEEE Computer, Vol. 17, No. 6, pp. 8 - 19. 17.2 Journal Articles and Study Reports Brown, J.F., T.R. Loveland, J.W. Merchant, B.C. Reed, and D.O. Ohlen. 1993. Using Multisource Data in Global Land Characterization: Concepts, Requirements, and Methods. Photogrammetric Engineering and Remote Sensing, 59(6): 977-987. Eidenshink, J.C. 1992a. The 1990 Conterminous U.S. AVHRR Data Set. Photogrammetric Engineering and Remote Sensing 58: 809-813. Eidenshink, J.C. 1992b. The 1992 Conterminous U.S. AVHRR Biweekly Composites. Documentation File on 1992 1-km AVHRR Biweekly Image Composites on CD-ROM. Eidenshink, J.C. and J.L. Faundeen. 1994. The 1-km AVHRR global land data set: first stages in implementation, Int. J. of Rem. Sens., 15 (17): 3443-3462. Holben, B. N., Y.J. Kaufman, and J.D. Kendall. 1990. NOAA-11 AVHRR visible and near-IR inflight calibration. The International Journal of Remote Sensing, 11: 1511-1519. Loveland, T.R., J.W. Merchant, D.O. Ohlen, and J.F. Brown. 1991. Development of a Land Cover Characteristics Data Base for the Conterminous U.S. Photogrammetric Engineering and Remote Sensing, 57(11): 1453-1463. Los, S.O., C.J. Tucker, and C.O. Justice. 1995. Documentation File for Normalized Difference Vegetation Index (NDVI) Data Set on ISLSCP No. 1 CD-ROM. NASA Goddard Space Flight Center. Newcomer, J.A. 1992. Documentation File for FIFE Level-1 Advanced Very High Resolution Radiometer (AVHRR) Data Sets on CD-ROM. NASA Goddard Space Flight Center. Price, J.C. 1987. Calibration of satellite radiometers and the comparison of vegetation indices. Remote Sensing of the Environment, 21: 15-27. Rao, N.C R. 1987. Pre-launch calibration of channels 1 and 2 of Advanced Very High Resolution Radiometer. NOAA Technical Report NESDIS 36, Satellite Research Laboratory, National Environmental Satellite, Data, and Information Service. Washington, D.C. Sellers, P. and F. Hall. 1994. Boreal Ecosystem-Atmosphere Study: Experiment Plan. Version 1994-3.0. NASA BOREAS Report (EXPLAN 94). Sellers, P., F. Hall, H. Margolis, B. Kelly, D. Baldocchi, G. den Hartog, J. Cihlar, M.G. Ryan, B. Goodison, P. Crill, K.J. Ranson, D. Lettenmaier, and D.E. Wickland. 1995. The boreal ecosystem-atmosphere study (BOREAS): an overview and early results from the 1994 field year. Bulletin of the American Meteorological Society. 76(9):1549-1577. Sellers, P., F. Hall, and K.F. Huemmrich. 1996. Boreal Ecosystem-Atmosphere Study: 1994 Operations. NASA BOREAS Report (OPS DOC 94). Sellers, P. and F. Hall. 1996. Boreal Ecosystem-Atmosphere Study: Experiment Plan. Version 1996-2.0. NASA BOREAS Report (EXPLAN 96). Sellers, P., F. Hall, and K.F. Huemmrich. 1997. Boreal Ecosystem-Atmosphere Study: 1996 Operations. NASA BOREAS Report (OPS DOC 96). Sellers, P.J., F.G. Hall, R.D. Kelly, A. Black, D. Baldocchi, J. Berry, M. Ryan, K.J. Ranson, P.M. Crill, D.P. Lettenmaier, H. Margolis, J. Cihlar, J. Newcomer, D. Fitzjarrald, P.G. Jarvis, S.T. Gower, D. Halliwell, D. Williams, B. Goodison, D.E. Wickland, and F.E. Guertin. 1997. BOREAS in 1997: Experiment Overview, Scientific Results and Future Directions. Journal of Geophysical Research. BOREAS Special Issue. 102(D24): 28731-28770. Steyaert, L.T., F.G. Hall, and T.R. Loveland. 1997. Land Cover Mapping, Fire Disturbance-Regeneration, and Multiresolution Land Cover Scaling Studies in the BOREAS Forest Ecosystem with Multiresolution 1-km AVHRR. J. Geophys. Res. 102: 29581-29598. Teillet, P.M., and Holben, B.N. 1991. unpublished report. Teillet, P.M., and Holben, B.N. 1994. Towards Operational Radiometric Calibration of NOAA-AVHRR Imagery in the Visible and Infrared Channels. Canadian Journal of Remote Sensing, 20(1): 1-10. 17.3 Archive/DBMS Usage Documentation None. 18. Glossary of Terms None. 19. List of Acronyms AEAC - Albers Equal Area Conic AFM - Airborne Fluxes and Meteorology APT - Automatic Picture Transmission ASCII - American Standard Code for Information Interchange BOREAS - Boreal Ecosystem-Atmosphere Study BORIS - BOREAS Information System BPI - Bytes Per Inch CCRS - Canadian Centre for Remote Sensing CD-ROM - Compact Disk-Read-Only-Memory DAAC - Distributed Active Archive Center DAT - Digital Archive Tape DEM - Digital Elevation Model EDC - EROS Data Center EOS - Earth Observing System EOSAT - Earth Observing Satellite Company EOSDIS - EOS Data and Information System EROS - Earth Resources Observation System GAC - Global Area Coverage GCM - Global Circulation Model GMT - Greenwich Mean Time GPS - Global Positioning System GRS80 - Geodetic Reference System of 1980 GSFC - Goddard Space Flight Center HRPT - Higher Resolution Picture Transmission IFC - Intensive Field Campaign IFOV - Instantaneous Field of View LAC - Local Area Coverage LST - Local Standard Time MSA - Modeling Sub-Area NAD27 - North American Datum 1927 NAD83 - North American Datum 1983 NASA - National Aeronautics and Space Administration NeDT - Noise Equivalent Differential Temperature NDVI - Normalized Difference Vegetation Index NOAA - National Oceanic and Atmospheric Administration NRL - Naval Research Laboratory NSA - Northern Study Area ORNL - Oak Ridge National Laboratory PANP - Prince Albert National Park RSS - Remote Sensing Science SSA - Southern Study Area SST - Sea Surface Temperature SVAT - Surface Vegetation and Atmosphere TE - Terrestrial Ecology TF - Tower Fluxes TGB - Trace Gas Biogeochemistry TIROS - Television and Infrared Observation Satellite TM - Thematic Mapper URL - Uniform Resource Locator USGS - United States Geological Survey UTM - Universal Transverse Mercator WGS84 - World Geodetic System of 1984 WRS - Worldwide Reference System WWW - World Wide Web 20. Document Information 20.1 Document Revision Date Written: 01-Oct-1996 Last Updated: 24-Nov-1998 20.2 Document Review Date(s) BORIS Review: 11-Sep-1998 Science Review: 20.3 Document ID 20.4 Citation These data were classified as a part of investigation TE-18, PI F.G. Hall, using Landsat 5 TM data from 06-Aug-1990. The BOREAS science staff produced this data product. Any publication of these data should acknowledge the source of the data as the Canadian Centre for Remote Sensing (CCRS). 20.5 Document Curator 20.6 Document URL Keywords AVHRR LAND COVER CLASSIFICATION AFM12_AVHRR_CLASS.doc 03/03/99