6 inch resolution raster image of New York City, classified by landcover type. High resolution land cover data set for New York City. This is the 6 inch version of the high-resolution land cover dataset for New York City. Seven land cover classes were mapped: (1) tree canopy, (2) grass/shrub, (3) bare earth, (4) water, (5) buildings, (6) roads, and (7) other paved surfaces. The minimum mapping unit for the delineation of features was set at 3 square feet. The primary sources used to derive this land cover layer were the 2010 LiDAR and the 2008 4-band orthoimagery. Ancillary data sources included GIS data (city boundary, building footprints, water, parking lots, roads, railroads, railroad structures, ballfields) provided by New York City (all ancillary datasets except railroads); UVM Spatial Analysis Laboratory manually created railroad polygons from manual interpretation of 2008 4-band orthoimagery. The tree canopy class was considered current as of 2010; the remaining land-cover classes were considered current as of 2008. Object-Based Image Analysis (OBIA) techniques were employed to extract land cover information using the best available remotely sensed and vector GIS datasets. OBIA systems work by grouping pixels into meaningful objects based on their spectral and spatial properties, while taking into account boundaries imposed by existing vector datasets. Within the OBIA environment a rule-based expert system was designed to effectively mimic the process of manual image analysis by incorporating the elements of image interpretation (color/tone, texture, pattern, location, size, and shape) into the classification process. A series of morphological procedures were employed to insure that the end product is both accurate and cartographically pleasing. More than 35,000 corrections were made to the classification. Overall accuracy was 96%. This dataset was developed as part of the Urban Tree Canopy (UTC) Assessment for New York City. As such, it represents a 'top down' mapping perspective in which tree canopy over hanging other features is assigned to the tree canopy class. At the time of its creation this dataset represents the most detailed and accurate land cover dataset for the area. This project was funded by National Urban and Community Forestry Advisory Council (NUCFAC) and the National Science Fundation (NSF), although it is not specifically endorsed by either agency. The methods used were developed by the University of Vermont Spatial Analysis Laboratory, in collaboration with the New York City Urban Field Station, with funding from the USDA Forest Service.

Abstract

The Urban Green Raster Germany is a land cover classification for Germany that addresses in particular the urban vegetation areas. The raster dataset covers the terrestrial national territory of Germany and has a spatial resolution of 10 meters. The dataset is based on a fully automated classification of Sentinel-2 satellite data from a full 2018 vegetation period using reference data from the European LUCAS land use and land cover point dataset.
The dataset identifies eight land cover classes. These include Built-up, Built-up with significant green share, Coniferous wood, Deciduous wood, Herbaceous vegetation (low perennial vegetation), Water, Open soil, Arable land (low seasonal vegetation).
The land cover dataset provided here is offered as an integer raster in GeoTiff format. The assignment of the number coding to the corresponding land cover class is explained in the legend file.

Data acquisition

The data acquisition comprises two main processing steps: (1) Collection, processing, and automated classification of the multispectral Sentinel 2 satellite data with the “Land Cover DE method”, resulting in the raw land cover classification dataset, NDVI layer, and RF assignment frequency vector raster. (2) GIS-based postprocessing including discrimination of (densely) built-up and loosely built-up pixels according NDVI threshold, and creating water-body and arable-land masks from geo-topographical base-data (ATKIS Basic DLM) and reclassification of water and arable land pixels based on the assignment frequency.

Data collection

Satellite data were searched and downloaded from the Copernicus Open Access Hub (https://scihub.copernicus.eu/).

The LUCAS reference and validation points were loaded from the Eurostat platform (https://ec.europa.eu/eurostat/web/lucas/data/database).

The processing of the satellite data was performed at the DLR data center in Oberpfaffenhofen.

GIS-based post-processing of the automatic classification result was performed at IOER in Dresden.

Value of the data

The dataset can be used to quantify the amount of green areas within cities on a homogeneous data base [5].

Thus it is possible to compare cities of different sizes regarding their greenery and with respect to their ratio of green and built-up areas [6].

Built-up areas within cities can be discriminated regarding their built-up density (dense built-up vs. built-up with higher green share).

Data description

A Raster dataset in GeoTIFF format: The dataset is stored as an 8 bit integer raster with values ranging from 1 to 8 for the eight different land cover classes. The nomenclature of the coded values is as follows: 1 = Built-up, 2=open soil; 3=Coniferous wood, 4= Deciduous wood, 5=Arable land (low seasonal vegetation), 6=Herbaceous vegetation (low perennial vegetation), 7=Water, 8=Built-up with significant green share. Name of the file ugr2018_germany.tif. The dataset is zipped alongside with accompanying files: *.twf (geo-referencing world-file), *.ovr (Overlay file for quick data preview in GIS), *.clr (Color map file).

A text file with the integer value assignment of the land cover classes. Name of the file: Legend_LC-classes.txt.

Experimental design, materials and methods

The first essential step to create the dataset is the automatic classification of a satellite image mosaic of all available Sentinel-2 images from May to September 2018 with a maximum cloud cover of 60 percent. Points from the 2018 LUCAS (Land use and land cover survey) dataset from Eurostat [1] were used as reference and validation data. Using Random Forest (RF) classifier [2], seven land use classes (Deciduous wood, Coniferous wood, Herbaceous vegetation (low perennial vegetation), Built-up, Open soil, Water, Arable land (low seasonal vegetation)) were first derived, which is methodologically in line with the procedure used to create the dataset "Land Cover DE - Sentinel-2 - Germany, 2015" [3]. The overall accuracy of the data is 93 % [4].

Two downstream post-processing steps served to further qualify the product. The first step included the selective verification of pixels of the classes arable land and water. These are often misidentified by the classifier due to radiometric similarities with other land covers; in particular, radiometric signatures of water surfaces often resemble shadows or asphalt surfaces. Due to the heterogeneous inner-city structures, pixels are also frequently misclassified as cropland.

To mitigate these errors, all pixels classified as water and arable land were matched with another data source. This consisted of binary land cover masks for these two land cover classes originating from the Monitor of Settlement and Open Space Development (IOER Monitor). For all water and cropland pixels that were outside of their respective masks, the frequencies of class assignments from the RF classifier were checked. If the assignment frequency to water or arable land was at least twice that to the subsequent class, the classification was preserved. Otherwise, the classification strength was considered too weak and the pixel was recoded to the land cover with the second largest assignment frequency.

Furthermore, an additional land cover class "Built-up with significant vegetation share" was introduced. For this purpose, all pixels of the Built-up class were intersected with the NDVI of the satellite image mosaic and assigned to the new category if an NDVI threshold was exceeded in the pixel. The associated NDVI threshold was previously determined using highest resolution reference data of urban green structures in the cities of Dresden, Leipzig and Potsdam, which were first used to determine the true green fractions within the 10m Sentinel pixels, and based on this to determine an NDVI value that could be used as an indicator of a significant green fraction within the built-up pixel. However, due to the wide dispersion of green fraction values within the built-up areas, it is not possible to establish a universally valid green percentage value for the land cover class of Built-up with significant vegetation share. Thus, the class essentially serves to the visual differentiability of densely and loosely (i.e., vegetation-dominated) built-up areas.

Acknowledgments

This work was supported by the Federal Institute for Research on Building, Urban Affairs and Spatial Development (BBSR) [10.06.03.18.101].The provided data has been developed and created in the framework of the research project “Wie grün sind bundesdeutsche Städte?- Fernerkundliche Erfassung und stadträumlich-funktionale Differenzierung der Grünausstattung von Städten in Deutschland (Erfassung der urbanen Grünausstattung)“ (How green are German cities?- Remote sensing and urban-functional differentiation of the green infrastructure of cities in Germany (Urban Green Infrastructure Inventory)). Further persons involved in the project were: Fabian Dosch (funding administrator at BBSR), Stefan Fina (research partner, group leader at ILS Dortmund), Annett Frick, Kathrin Wagner (research partners at LUP Potsdam).

References

[1] Eurostat (2021): Land cover / land use statistics database LUCAS. URL: https://ec.europa.eu/eurostat/web/lucas/data/database

[2] L. Breiman (2001). Random forests, Mach. Learn., 45, pp. 5-32

[3] M. Weigand, M. Wurm (2020). Land Cover DE - Sentinel-2—Germany, 2015 [Data set]. German Aerospace Center (DLR). doi: 10.15489/1CCMLAP3MN39

[4] M. Weigand, J. Staab, M. Wurm, H. Taubenböck, (2020). Spatial and semantic effects of LUCAS samples on fully automated land use/land cover classification in high-resolution Sentinel-2 data. Int J Appl Earth Obs, 88, 102065. doi: https://doi.org/10.1016/j.jag.2020.102065

[5] L. Eichler., T. Krüger, G. Meinel, G. (2020). Wie grün sind deutsche Städte? Indikatorgestützte fernerkundliche Erfassung des Stadtgrüns. AGIT Symposium 2020, 6, 306–315. doi: 10.14627/537698030

[6] H. Taubenböck, M. Reiter, F. Dosch, T. Leichtle, M. Weigand, M. Wurm (2021). Which city is the greenest? A multi-dimensional deconstruction of city rankings. Comput Environ Urban Syst, 89, 101687. doi: 10.1016/j.compenvurbsys.2021.101687

A 6-in resolution 8-class land cover dataset derived from the 2017 Light Detection and Ranging (LiDAR) data capture. This dataset was developed as part of an updated urban tree canopy assessment and therefore represents a ''top-down" mapping perspective in which tree canopy overhanging features is assigned to the tree canopy class. The eight land cover classes mapped were: (1) Tree Canopy, (2) Grass\Shrubs, (3) Bare Soil, (4) Water, (5) Buildings, (6) Roads, (7) Other Impervious, and (8) Railroads. The primary sources used to derive this land cover layer were 2017 LiDAR (1-ft post spacing) and 2016 4-band orthoimagery (0.5-ft resolution). Object based image analysis was used to automate land-cover features using LiDAR point clouds and derivatives, orthoimagery, and vector GIS datasets -- City Boundary (2017, NYC DoITT) Buildings (2017, NYC DoITT) Hydrography (2014, NYC DoITT) LiDAR Hydro Breaklines (2017, NYC DoITT) Transportation Structures (2014, NYC DoITT) Roadbed (2014, NYC DoITT) Road Centerlines (2014, NYC DoITT) Railroads (2014, NYC DoITT) Green Roofs (date unknown, NYC Parks) Parking Lots (2014, NYC DoITT) Parks (2016, NYC Parks) Sidewalks (2014, NYC DoITT) Synthetic Turf (2018, NYC Parks) Wetlands (2014, NYC Parks) Shoreline (2014, NYC DoITT) Plazas (2014, NYC DoITT) Utility Poles (2014, ConEdison via NYCEM) Athletic Facilities (2017, NYC Parks) For the purposes of classification, only vegetation > 8 ft were classed as Tree Canopy. Vegetation below 8 ft was classed as Grass/Shrub. To learn more about this dataset, visit the interactive "Understanding the 2017 New York City LiDAR Capture" Story Map -- https://maps.nyc.gov/lidar/2017/ Please see the following link for additional documentation on this dataset -- https://github.com/CityOfNewYork/nyc-geo-metadata/blob/master/Metadata/Metadata_LandCover.md

The Raster Based GIS Coverage of Mexican Population is a gridded coverage (1 x 1 km) of Mexican population. The data were converted from vector into raster. The population figures were derived based on available point data (the population of known localities - 30,000 in all). Cell values were derived using a weighted moving average function (Burrough, 1986), and then calculated based on known population by state. The result from this conversion is a coverage whose population data is based on square grid cells rather than a series of vectors. This data set is produced by the Columbia University Center for International Earth Science Information Network (CIESIN) in collaboration with the Instituto Nacional de Estadistica Geografia e Informatica (INEGI).

This dataset is called the Gridded SSURGO (gSSURGO) Database and is derived from the Soil Survey Geographic (SSURGO) Database. SSURGO is generally the most detailed level of soil geographic data developed by the National Cooperative Soil Survey (NCSS) in accordance with NCSS mapping standards. The tabular data represent the soil attributes, and are derived from properties and characteristics stored in the National Soil Information System (NASIS). The gSSURGO data were prepared by merging traditional SSURGO digital vector map and tabular data into a Conterminous US-wide extent, and adding a Conterminous US-wide gridded map layer derived from the vector, plus a new value added look up (valu) table containing "ready to map" attributes. The gridded map layer is offered in an ArcGIS file geodatabase raster format. The raster and vector map data have a Conterminous US-wide extent. The raster map data have a 30 meter cell size. Each cell (and polygon) is linked to a map unit identifier called the map unit key. A unique map unit key is used to link to raster cells and polygons to attribute tables, including the new value added look up (valu) table that contains additional derived data. The value added look up (valu) table contains attribute data summarized to the map unit level using best practice generalization methods intended to meet the needs of most users. The generalization methods include map unit component weighted averages and percent of the map unit meeting a given criteria. The Gridded SSURGO dataset was created for use in national, regional, and state-wide resource planning and analysis of soils data. The raster map layer data can be readily combined with other national, regional, and local raster layers, e.g., National Land Cover Database (NLCD), the National Agricultural Statistics Service (NASS) Crop Data Layer, or the National Elevation Dataset (NED).

scripts.zip

arcgisTools.atbx: terrainDerivatives: make terrain derivatives from digital terrain model (Band 1 = TPI (50 m radius circle), Band 2 = square root of slope, Band 3 = TPI (annulus), Band 4 = hillshade, Band 5 = multidirectional hillshades, Band 6 = slopeshade). rasterizeFeatures: convert vector polygons to raster masks (1 = feature, 0 = background).

makeChips.R: R function to break terrain derivatives and chips into image chips of a defined size. makeTerrainDerivatives.R: R function to generated 6-band terrain derivatives from digital terrain data (same as ArcGIS Pro tool). merge_logs.R: R script to merge training logs into a single file. predictToExtents.ipynb: Python notebook to use trained model to predict to new data. trainExperiments.ipynb: Python notebook used to train semantic segmentation models using PyTorch and the Segmentation Models package. assessmentExperiments.ipynb: Python code to generate assessment metrics using PyTorch and the torchmetrics library. graphs_results.R: R code to make graphs with ggplot2 to summarize results. makeChipsList.R: R code to generate lists of chips in a directory. makeMasks.R: R function to make raster masks from vector data (same as rasterizeFeatures ArcGIS Pro tool).

terraceDL.zip

dems: LiDAR DTM data partitioned into training, testing, and validation datasets based on HUC8 watershed boundaries. Original DTM data were provided by the Iowa BMP mapping project: https://www.gis.iastate.edu/BMPs. extents: extents of the training, testing, and validation areas as defined by HUC 8 watershed boundaries. vectors: vector features representing agricultural terraces and partitioned into separate training, testing, and validation datasets. Original digitized features were provided by the Iowa BMP Mapping Project: https://www.gis.iastate.edu/BMPs.

This packaged data collection contains all of the outputs from our primary model, including the following data layers: Habitat Cores (vector polygons) Least-cost Paths (vector lines) Least-cost Corridors (raster) Least-cost Corridors (vector polygon interpretation) Modeling Extent (vector polygon) Please refer to the embedded spatial metadata and the information in our full report for details on the development of these data layers. Packaged data are available in two formats: Geodatabase (.gdb): A related set of file geodatabase rasters and feature classes, packaged in an ESRI file geodatabase. ArcGIS Pro Map Package (.mpkx): The same data included in the geodatabase, presented as fully-symbolized layers in a map. Note that you must have ArcGIS Pro version 2.0 or greater to view. See Cross-References for links to individual datasets, which can be downloaded in shapefile (.shp) or raster GeoTIFF (.tif) formats.

Abstract

This dataset and its metadata statement were supplied to the Bioregional Assessment Programme by a third party and are presented here as originally supplied.

This dataset is a complete state-wide digital land use map of Queensland. The dataset is a product of the Queensland Land Use Mapping Program (QLUMP) and was produced by the Queensland Government. It presents the most current mapping of land use features for Queensland, including the land use mapping products from 1999, 2006 and 2009, in a single feature layer. This dataset was last updated July 2012. See additional information also.

Purpose

Indicates the current primary use or management objective of the land.

Dataset History

Source DataQueensland Government - Land use mapping (1999); Landsat TM and ETM imagery; Spot5 imagery; High resolution ortho photography through the Spatial Imagery Subscription Plan (SISP); Queensland Digital Cadastral Database (DCDB) (2009), Queensland Valuation and Sales Database (QVAS) (2009); Queensland Nature Refuges (2009); Queensland Estates (2009); Queensland Herbarium's Regional Ecosystem, Water Body and Wetlands datasets (2009); Statewide Landcover & Trees Study (SLATS) Queensland Dams and Waterbodies (2009) and land cover change data; scanned aerial photography (1999-2009).Additional verbal & written information on land uses & their locations was obtained from regional Queensland Government officers, Local Government Authorities, land owners & managers, private industry as well as from field observations & checking.Data captureA range of existing digital datasets containing land use information was collated from the Queensland Government spatial data inventory and prepared for use in a GIS using ArcGIS and ERDAS Imagine software.Processing steps To compile the 1999 baseline mapping, datasets containing baseline land cover (supplied by SLATS), Protected Areas, State Forest and Timber Reserves, plantations, coastal wetlands, reserves (from DCDB) and logged forests were interpreted in a spatial model to produce a preliminary land use raster image.The model incorporated a decision matrix which assigned each pixel a specific land use class according to a set of pre-determined rules.Individual catchments were clipped from the model output and enhanced with additional land use information interpreted primarily from Landsat TM and ETM imagery as well as scanned and hardcopy aerial photography (where available). The DCDB and other datasets containing land use information were used to help identify property and land use type boundaries. This process produced a draft land use raster.Verification of the draft land use dataset, particularly those with significant areas of intensive land uses, was undertaken by comparing mapped land use classes with observed land use classes in the field where possible. The final raster image was converted to a vector coverage in ARC/Info and GIS editing performed.The existing 1999 baseline (or later where available) land use dataset (vector) formed the basis for the 2006 and 2009 land use mapping. The 2006 & 2009 datasets were then updated primarily by interpretation of SPOT5 imagery, high-res orthophotography, scanned aerial photography and inclusion of expert local knowledge. This was performed in an ESRI ArcSDE geodatabase replication infrastructure, across some nine regional offices. The DCDB, QVAS, Estates, Queensland Herbarium wetlands and SLATS land cover change and waterbody datasets were used to assist in identification and delineation of property and land use type boundaries. Digitised areas of uniform land use type were assigned to land use classes according to ALUMC Version 7 (May 2010).This "current" land use mapping product presents a complete state-wide land use map of Queensland, after collating the most current land use datasets within a single mapping layer.An independent validation was undertaken to assess thematic (attribute) accuracy under the ALUM classification. Please refer to the orignal source data for the validation results.

Dataset Citation

Queensland Department of Science, Information Technology, Innovation and the Arts (2013) Bioregional_Assessment_Programme_Land use mapping - Queensland current. Bioregional Assessment Source Dataset. Viewed 21 December 2017, http://data.bioregionalassessments.gov.au/dataset/740d257f-b622-49c2-9745-be283239add3.

Wind Farms - follows on from the 'Dave' Data Download case study. View and symbolise OS raster and height data and Wind Farm location data. GIS vector data. This dataset was first accessioned in the EDINA ShareGeo Open repository on 2014-04-10 and migrated to Edinburgh DataShare on 2017-02-22.

Abstract

This dataset and its metadata statement were supplied to the Bioregional Assessment Programme by a third party and are presented here as originally supplied.

Resource contains an ArcGIS file geodatabase raster for the National Vegetation Information System (NVIS) Major Vegetation Groups - Australia-wide, present extent (FGDB_NVIS4_1_AUST_MVG_EXT).

Related datasets are also included: FGDB_NVIS4_1_KEY_LAYERS_EXT - ArcGIS File Geodatabase Feature Class of the Key Datasets that make up NVIS Version 4.1 - Australia wide; and FGDB_NVIS4_1_LUT_KEY_LAYERS - Lookup table for Dataset Key Layers.

This raster dataset provides the latest summary information (November 2012) on Australia's present (extant) native vegetation. It is in Albers Equal Area projection with a 100 m x 100 m (1 Ha) cell size. A comparable Estimated Pre-1750 (pre-european, pre-clearing) raster dataset is available: - NVIS4_1_AUST_MVG_PRE_ALB. State and Territory vegetation mapping agencies supplied a new version of the National Vegetation Information System (NVIS) in 2009-2011. Some agencies did not supply new data for this version but approved re-use of Version 3.1 data. Summaries were derived from the best available data in the NVIS extant theme as at June 2012. This product is derived from a compilation of data collected at different scales on different dates by different organisations. Please refer to the separate key map showing scales of the input datasets. Gaps in the NVIS database were filled by non-NVIS data, notably parts of South Australia and small areas of New South Wales such as the Curlewis area. The data represent on-ground dates of up to 2006 in Queensland, 2001 to 2005 in South Australia (depending on the region) and 2004/5 in other jurisdictions, except NSW. NVIS data was partially updated in NSW with 2001-09 data, with extensive areas of 1997 data remaining from the earlier version of NVIS. Major Vegetation Groups were identified to summarise the type and distribution of Australia's native vegetation. The classification contains different mixes of plant species within the canopy, shrub or ground layers, but are structurally similar and are often dominated by a single genus. In a mapping sense, the groups reflect the dominant vegetation occurring in a map unit where there are a mix of several vegetation types. Subdominant vegetation groups which may also be present in the map unit are not shown. For example, the dominant vegetation in an area may be mapped as dominated by eucalypt open forest, although it contains pockets of rainforest, shrubland and grassland vegetation as subdominants. The (related) Major Vegetation Subgroups represent more detail about the understorey and floristics of the Major Vegetation Groups and are available as separate raster datasets: - NVIS4_1_AUST_MVS_EXT_ALB - NVIS4_1_AUST_MVS_PRE_ALB A number of other non-vegetation and non-native vegetation land cover types are also represented as Major Vegetation Groups. These are provided for cartographic purposes, but should not be used for analyses. For further background and other NVIS products, please see the links on http://www.environment.gov.au/erin/nvis/index.html.

The current NVIS data products are available from http://www.environment.gov.au/land/native-vegetation/national-vegetation-information-system.

Purpose

For use in Bioregional Assessment land classification analyses

Dataset History

NVIS Version 4.1

The input vegetation data were provided from over 100 individual projects representing the majority of Australia's regional vegetation mapping over the last 50 years. State and Territory custodians translated the vegetation descriptions from these datasets into a common attribute framework, the National Vegetation Information System (ESCAVI, 2003). Scales of input mapping ranged from 1:25,000 to 1:5,000,000. These were combined into an Australia-wide set of vector data. Non-terrestrial areas were mostly removed by the State and Territory custodians before supplying the data to the Environmental Resources Information Network (ERIN), Department of Sustainability Environment Water Population and Communities (DSEWPaC).

Each NVIS vegetation description was written to the NVIS XML format file by the custodian, transferred to ERIN and loaded into the NVIS database at ERIN. A considerable number of quality checks were performed automatically by this system to ensure conformity to the NVIS attribute standards (ESCAVI, 2003) and consistency between levels of the NVIS Information Hierarchy within each description. Descriptions for non-vegetation and non-native vegetation mapping codes were transferred via CSV files.

The NVIS vector (polygon) data for Australia comprised a series of jig-saw pieces, eachup to approx 500,000 polygons - the maximum tractable size for routine geoprocesssing. The spatial data was processed to conform to the NVIS spatial format (ESCAVI, 2003; other papers). Spatial processing and attribute additions were done mostly in ESRI File Geodatabases. Topology and minor geometric corrections were also performed at this stage. These datasets were then loaded into ESRI Spatial Database Engine as per the ERIN standard. NVIS attributes were then populated using Oracle database tables provided by custodians, mostly using PL/SQL Developer or in ArcGIS using the field calculator (where simple).

Each spatial dataset was joined to and checked against a lookup table for the relevant State/Territory to ensure that all mapping codes in the dominant vegetation type of each polygon (NVISDSC1) had a valid lookup description, including an allocated MVG. Minor vegetation components of each map unit (NVISDSC2-6) were not checked, but could be considered mostly complete.

Each NVIS vegetation description was allocated to a Major Vegetation Group (MVG) by manual interpretation at ERIN. The Australian Natural Resources Atlas (http://www.anra.gov.au/topics/vegetation/pubs/native_vegetation/vegfsheet.html) provides detailed descriptions of most Major Vegetation Groups. Three new MVGs were created for version 4.1 to better represent open woodland formations and forests (in the NT) with no further data available. NVIS vegetation descriptions were reallocated into these classes, if appropriate:

• Unclassified Forest

• Other Open Woodlands

• Mallee Open Woodlands and Sparse Mallee Shublands

(Thus there are a total of 33 MVGs existing as at June 2012). Data values defined as cleared or non-native by data custodians were attributed specific MVG values such as 25 - Cleared or non native, 27 - naturally bare, 28 - seas & estuaries, and 99 - Unknown.

As part of the process to fill gaps in NVIS, the descriptive data from non-NVIS sources was also referenced in the NVIS database, but with blank vegetation descriptions. In general. the gap-fill data comprised (a) fine scale (1:250K or better) State/Territory vegetation maps for which NVIS descriptions were unavailable and (b) coarse-scale (1:1M) maps from Commonwealth and other sources. MVGs were then allocated to each description from the available desciptions in accompanying publications and other sources.

Parts of New South Wales, South Australia, QLD and the ACT have extensive areas of vector "NoData", thus appearing as an inland sea. The No Data areas were dealt with differently by state. In the ACT and SA, the vector data was 'gap-filled' and attributed using satellite imagery as a guide prior to rasterising. Most of these areas comprised a mixture of MVG 24 (inland water) and 25 (cleared), and in some case 99 (Unknown). The NSW & QLD 'No Data' areas were filled using a raster mask to fill the 'holes'. These areas were attributed with MVG 24, 26 (water & unclassified veg), MVG 25 (cleared); or MVG 99 Unknown/no data, where these areas were a mixture of unknown proportions.

Each spatial dataset with joined lookup table (including MVG_NUMBER linked to NVISDSC1) was exported to a File Geodatabase as a feature class. These were reprojected into Albers Equal Area projection (Central_Meridian: 132.000000, Standard_Parallel_1: -18.000000, Standard_Parallel_2: -36.000000, Linear Unit: Meter (1.000000), Datum GDA94, other parameters 0).

Each feature class was then rasterised to a 100m raster with extents to a multiple of 1000 m, to ensure alignment. In some instances, areas of 'NoData' had to be modelled in raster. For example, in NSW where non-native areas (cleared, water bodies etc) have not been mapped. The rasters were then merged into a 'state wide' raster. State rasters were then merged into this 'Australia wide' raster dataset.

November 2012 Corrections

Closer inspection of the original 4.1 MVG Extant raster dataset highlighted some issues with the raster creation process which meant that raster pixels in some areas did not align as intended. These were corrected, and the new properly aligned rasters released in November 2012.

Dataset Citation

Department of the Environment (2012) Australia - Present Major Vegetation Groups - NVIS Version 4.1 (Albers 100m analysis product). Bioregional Assessment Source Dataset. Viewed 10 July 2017, http://data.bioregionalassessments.gov.au/dataset/57c8ee5c-43e5-4e9c-9e41-fd5012536374.

The Hills of Governor's Island Dataset for GRASS GIS This geospatial dataset contains raster and vector data for the Hills region of Governor's Island, New York City, USA. The top level directory governors_island_hills_for_grass is a GRASS GIS location for NAD_1983_StatePlane_New_York_Long_Island_FIPS_3104_Feet in US Surveyor's Feet with EPSG code 2263. Inside the location there is the PERMANENT mapset, a license file, data record, readme file, workspace, color table, category rules, and scripts for data processing. This dataset was created for the course GIS for Designers.

Instructions Install GRASS GIS, unzip this archive, and move the location into your GRASS GIS database directory. If you are new to GRASS GIS read the first time users guide.

Data Sources

https://orthos.dhses.ny.gov/

https://data.cityofnewyork.us/

Maps

Orthophotographs from 2012, 2014, 2016, 2018, and 2020

Digital elevation model from 2017

Digital surface models from 2014 and 2017

Landcover from 2014

License This dataset is licensed under the ODC Public Domain Dedication and License 1.0 (PDDL) by Brendan Harmon.

Basic Global Dataset for GRASS GIS This geospatial dataset contains global raster and vector data. The top level directory global-dataset is a GRASS GIS location for the World Geodetic System 1984 (WGS84) with EPSG code 4326. Inside the location there is the PERMANENT mapset, a license file, and readme file. Instructions Install GRASS GIS, unzip this archive, and move the location into your GRASS GIS database directory. If you are new to GRASS GIS read the first time users guide. Data Source Natural Earth License This dataset is licensed under the ODC Public Domain Dedication and License 1.0 (PDDL) by Brendan Harmon.

New Orleans Dataset for GRASS GIS
This geospatial dataset contains raster and vector data for New Orleans, Louisiana, USA. The top level directory new-orleans-dataset is a GRASS GIS location for the North American Datum of 1983 (NAD 83) / Louisiana South State Plane Feet with EPSG code 3452. Inside the location there are the PERMANENT mapset with citywide data, a vieux_carre mapset with data for the French Quarter, Python scripts for data processing, data records, a color table, a license file, and readme file.

Instructions
Install GRASS GIS, unzip this archive, and move the location into your GRASS GIS database directory. If you are new to GRASS GIS read the first time users guide.

Data Sources

• U.S. Army Corps of Engineers 2012 Lidar Survey of New Orleans
• New Orleans Open Data

License
This dataset is licensed under the ODC Public Domain Dedication and License 1.0 (PDDL) by Brendan Harmon. The scripts are licensed under the GNU General Public License 3.0 by Brendan Harmon. The graphics are licensed under the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0) by Brendan Harmon.

High resolution land cover dataset for City of Boston, MA. Seven land cover classes were mapped: (1) tree canopy, (2) grass/shrub, (3) bare earth, (4) water, (5) buildings, (6) roads, and (7) other paved surfaces. The primary sources used to derive this land cover layer were 2013 LiDAR data, 2014 Orthoimagery, and 2016 NAIP imagery. Ancillary data sources included GIS data provided by City of Boston, MA or created by the UVM Spatial Analysis Laboratory. Object-based image analysis techniques (OBIA) were employed to extract land cover information using the best available remotely sensed and vector GIS datasets. OBIA systems work by grouping pixels into meaningful objects based on their spectral and spatial properties, while taking into account boundaries imposed by existing vector datasets. Within the OBIA environment a rule-based expert system was designed to effectively mimic the process of manual image analysis by incorporating the elements of image interpretation (color/tone, texture, pattern, location, size, and shape) into the classification process. A series of morphological procedures were employed to insure that the end product is both accurate and cartographically pleasing. Following the automated OBIA mapping a detailed manual review of the dataset was carried out at a scale of 1:2500 and all observable errors were corrected.

High resolution land cover dataset for City of Boston, MA. Seven land cover classes were mapped: (1) tree canopy, (2) grass/shrub, (3) bare earth, (4) water, (5) buildings, (6) roads, and (7) other paved surfaces. The primary sources used to derive this land cover layer were 2013 LiDAR data, 2014 Orthoimagery, and 2016 NAIP imagery. Ancillary data sources included GIS data provided by City of Boston, MA or created by the UVM Spatial Analysis Laboratory. Object-based image analysis techniques (OBIA) were employed to extract land cover information using the best available remotely sensed and vector GIS datasets. OBIA systems work by grouping pixels into meaningful objects based on their spectral and spatial properties, while taking into account boundaries imposed by existing vector datasets. Within the OBIA environment a rule-based expert system was designed to effectively mimic the process of manual image analysis by incorporating the elements of image interpretation (color/tone, texture, pattern, location, size, and shape) into the classification process. A series of morphological procedures were employed to insure that the end product is both accurate and cartographically pleasing. Following the automated OBIA mapping a detailed manual review of the dataset was carried out at a scale of 1:2500 and all observable errors were corrected.

Credits: University of Vermont Spatial Analysis Laboratory in collaboration with the City of Boston, Trust for Public Lands, and City of Cambridge.

New Orleans Dataset for ArcGIS
This geodatabase contains citywide raster and vector data for New Orleans, Louisiana, USA in the North American Datum of 1983 (NAD 83) / Louisiana South State Plane Feet with EPSG code 3452. Unzip the archive and open in ArcGIS.

Data Sources

• U.S. Army Corps of Engineers 2012 Lidar Survey of New Orleans
• New Orleans Open Data

License
This dataset is licensed under the ODC Public Domain Dedication and License 1.0 (PDDL) by Brendan Harmon.

One important reason for performing GIS analysis is to determine proximity. Often, this type of analysis is done using vector data and possibly the Buffer or Near tools. In this course, you will learn how to calculate distance using raster datasets as inputs in order to assign cells a value based on distance to the nearest source (e.g., city, campground). You will also learn how to allocate cells to a particular source and to determine the compass direction from a cell in a raster to a source.What if you don't want to just measure the straight line from one place to another? What if you need to determine the best route to a destination, taking speed limits, slope, terrain, and road conditions into consideration? In cases like this, you could use the cost distance tools in order to assign a cost (such as time) to each raster cell based on factors like slope and speed limit. From these calculations, you could create a least-cost path from one place to another. Because these tools account for variables that could affect travel, they can help you determine that the shortest path may not always be the best path.After completing this course, you will be able to:Create straight-line distance, direction, and allocation surfaces.Determine when to use Euclidean and weighted distance tools.Perform a least-cost path analysis.

Governor's Island Dataset for ArcGIS This archive contains an ArcGIS Pro project with a geodatabase of raster and vector data for Governor's Island, New York City, USA. The SRS is NAD83 / New York Long Island (ftUS) with the EPSG code 2263.

We developed a completely new free geospatial dataset and substituted all Spearfish (SD) examples in the previous editions with this new, much richer North Carolina (NC, USA) data set. This data set is a comprehensive collection of raster, vector and imagery data covering parts of North Carolina (NC), USA (map), prepared from public data sources provided by the North Carolina state and local government agencies and Global Land Cover Facility (GLCF).

This data is packaged as a GRASS location as well as SHAPE/GeoTIFF/KML/ArcGRID files. See also http://www.grassbook.org/data_menu3rd.php for download.

Basic Global Dataset for GRASS GIS
This geospatial dataset contains global raster and vector data from Natural Earth. The top level directory global-dataset is a GRASS GIS location for the World Geodetic System 1984 (WGS84) with EPSG code 4326. Inside the location there is the PERMANENT mapset, a license file, and readme file.

Instructions
Install GRASS GIS, extract this archive, and move the location into your GRASS GIS database directory. If you are new to GRASS GIS read the first time users guide.

Data Source

• Natural Earth

License
This dataset is licensed under the ODC Public Domain Dedication and License 1.0 (PDDL) by Brendan Harmon.

Statewide Ecopia 3 foot Land Cover (2021-2022)This raster land cover data is based off of high-resolution statewide imagery from 2021-2022. It was used by Ecopia to extract and digitize the entire state into 7 different land cover classes. Download Notes:This service can be entered into ArcGIS Pro where "Download Rasters" can be used to download approximately 20 square miles at a time. (Rt. click layer in TOC > Data > Download Rasters)Alternatively, the entire statewide 3ft dataset is available as a zipped download from here (includes colormap file): Ecopia_Statewide_3ft_Raster_TilesClasses available at bottom of this pages.Data SpecificationImagery Used for Extraction: Pixel resolution: 15 cm (6")Camera sensor: Hexagon Pushbroom (Content Mapper)Date of capture: 06/25/2021 - 08/14/2022Date of Vector Extraction: June 2023Extraction Methodology:Ecopia uses proprietary extraction and modeling software to process raw images into high-resolution land cover classifications.Quality Measurements:Measure Name - Threshold across Impervious Polygons:False Negatives <= 5% All PolygonsFalse Positives <= 5% All PolygonsValid Interpretation >= 95% All PolygonsMinimum Area 100% All PolygonsValid Geometry 100% All PolygonsMeasure Name - Threshold across Natural Polygons:False Negatives <=5% All PolygonsFalse Positives <=5% All PolygonsValid Interpretation >=90% All PolygonsMinimum Area 100% All PolygonsValid Geometry 100% All PolygonsLand Cover Classes:UnclassifiedImperviousImpervious, covered by treesShrub/low vegetationTree/forest/high vegetationOpen waterRailroadVegetation (Canopy Mapping)Tree canopy will be captured as a unique polygon layer. It can therefore overlap impervious layers.High vegetation is distinguished from low vegetation based on crown, texture, and derived height models. Leveraging stereo imagery produces results using 3D elevation models used to aid the distinction of vegetation categories. Distinguishing low from high vegetation is based on a 5m threshold, but this is not always feasible, especially in areas where heavy canopy prevents a visualization of the ground. In these circumstances, high vegetation will be given the priority over low vegetation. For more information visit: www.ecopiatech.comClasses:0: No data - Null, clear1: Unclassified2: Impervious3: Impervious, Covered by Tree Canopy6: Shrub/Low Vegetation7: Tree/Forest/High Vegetation8: Open Water12: Railroad