Cover-related Remote Sensing Data Products

Disentangling the Cover-related Concepts

   Cover data products provide measures of the proportions of a surface area covered/occupied by distinct ‘classes of cover’. Depending on the ‘classes of cover’ considered and the particular ‘layer’ of the surface area of interest different cover concepts are used, which can be a little bit confusing. These include Land Cover, Fractional Cover, and Ground Cover. Moreover, multiple varied classifications exist for some of these cover concepts, particularly Land Cover.

   All of these cover concepts can be estimated from remotely sensed imagery. Partially and temporally explicit data derived from these images have substantial applications in the assessment, monitoring, and management of ecosystem state and services, as well as agricultural productivity.

   Let us examine in more detail these Cover Concepts and other related concepts. For more information, see the Australian Bureau of Agriculture and Resource Economics and Sciences (ABARES) and the Food Agriculture Organization of the United Nations (FAO) links in the reference section.

Land Cover

   Refers to the biophysical and physical cover on the surface of the earth. Strictly speaking, it only includes biological and man-made features. However, in practice, the actual land (i.e. soil and rock types) and water features are also accounted as land cover. That is, typically a land cover classification will include vegetation types (trees, grasses,…), soil and rock types, anthropogenic elements (agriculture, built environments, asphalt,…), and water types (wetlands, open water,….).

Land Use

   Refers to the use to which the land cover is committed. That is, land use makes reference to how people use the landscape. Thus, land use combines land cover and the actions (activities, transformations, inputs,….) of people in it. Typically, land uses cannot exclusively be determined from remotely sensed imagery.

    Land cover and land use are often (and erroneously) used interchangeably. One of the reasons for these is that certain land uses, such as agriculture or urban, are associated with characteristic land cover categories. However, other land uses, such as nature conservation, do not have a one-on-one relationship with a particular land cover type. For example the land cover category ‘woodland’ can correspond to multiple land use categories, including ‘nature conservation’, ‘recreation’, ‘grazing’, and ‘timber production’. Conversely, the land use category ‘recreation’ could be applicable to different land cover categories, such as ‘woodland’, ‘lake’, or ‘urban’. The confusion between these two concepts is reviewed in Fisher et al. (2005).

Fractional Cover

   Fractional Cover typically refers to the fraction of an area, usually a pixel in the remote sensing context, that is covered by each of three specific types of cover: photosynthetic vegetation, non-photosynthetic vegetation (e.g.  senescent vegetation, leaf litter, and stubble), and bare landscape (e.g. soil, rocks, and recently burn areas). These three fractions are commonly and respectively referred as ‘green cover’, ‘brown cover’, and ‘bare cover’ fractions. In this type of remote sensing products is also common for the green and brown fraction to appear combined in what is known as the ‘total cover’ fraction. Sometimes fractional cover data products do not include all fractions, but only a subset of them.

Ground Cover

   Refers to the vegetation, biological crusts and stone that are in contact with the soil surface. Vegetation includes both green (i.e. photosynthetic vegetation) and non-green (i.e. non-photosynthetic vegetation and plant litter). Ground cover is thus a sub-component of land cover. From a remote sensing perspective, it is the fractional cover of the non-woody understorey. For this reason, ground cover is also referred to as Fractional Ground Cover.

    Ground cover patterns results from interactions between abiotic and biotic landscape attributes (topography, soil type, and vegetation dynamics), climate, and land management practices. For examples, high soil fertility and consistently high annual rainfall lead to naturally higher levels of ground cover, which could in turn be modified by land management practices..

Other Cover data products

   In some occasions, cover data products include different categories to those present in fractional cover or ground cover data products (i.e. bare cover, green cover, brown cover, and/or total cover; see above). For example, they can include categories such as ‘woody green cover’, ‘foliage projective cover’, or ‘water cover’ instead.

Uses of Cover Data in Ecosystem Management

   As mentioned above, cover data products have substantial applications in the management of ecosystems. Maps of cover status can assist on the assessment of the state of the landscape at a particular point in time. Time series of images/maps of cover are invaluable tools in the monitoring, modelling, and management (including evaluating past management decision) of natural resources. Ecosystem management applications of cover data include:

• Predicting and preventing wind and water soil erosion by estimating the green vegetation fraction/ground cover, which is intrinsically related to many soil processes such as infiltration, runoff and surface erosion.
• Fire prediction and management by estimating fuel load, which is in turn related to the type and amount of cover.
• Predicting and planning stocking rates in a grazing area by examining the ground cover.
• Conservation of natural resources by mapping fractional cover changes in space and time.
• Aid in the study of ecosystem processes such a nutrient cycling.

A Brief Note on Satellite Remote Sensing and TERN's Cover Data

   Currently TERN remotely sensed satellite data comes from two main sources: Landsat and MODIS. There are currently two active Landsat satellites, Landsats 7 and 8. Similarly, the Moderate Resolution Imaging Spectroradiometer (MODIS) imaging sensor is currently on board two satellites, Terra and Aqua. All four satellites (Landsats 7 and 8, Terra, and Aqua) are part of NASA’s Earth Observing System (EOS) program. The EOS program comprises (so far) 30 satellite missions for the long-term global observation of the land surface, oceans, biosphere, and atmosphere. A single data product for New South Wales was acquired using the SPOT-5 satellite. TERN’s satellite RS sources are briefly reviewed below in another Background Knowledge entry within the Landscape Assessment - Remote Sensing section of TERN's Data Skills Development Program (for further details see https://landsat.gsfc.nasa.gov, https://landsat.usgs.gov. and https://en.wikipedia.org).

   For ecological applications, the most relevant differences between Landsat and MODIS data are their spatial and temporal resolutions (see below). Landsat spectral bands have a moderate spatial resolution (15-100m) and low temporal resolution (8 days for the sensors in Landsat 7 and 8 combined). MODIS spectral bands, on the other hand have a low spatial resolution (250m - 1km) and a high temporal resolution (1 to 2 days for the sensors on board Terra and Aqua combined).  Consequently, MODIS is better suited to monitor larger-scale dynamics more frequently. Landsat, on the other hand, provides more spatial detail, but less frequently.

TERN RS Cover Data Products

   Numerous cover data products for Australia and Australian States/Regions are available from TERN data discovery portal. TERN range of cover data products include Fractional Cover, Ground Cover, Land Cover, and ‘Other Cover’ products (see above; Table 1). The majority of the Fractional Cover products include the three ‘traditional’ fractions in this type of data product: green, brown, and bare (see above). However, some of the fractional cover data products only include a single fraction or a combination of fractions, such as including only the ‘green cover fraction’ and/or the ‘total cover fraction’ (i.e. a combination of the green and brown cover fractions; see above).

   The vast majority of cover data products in TERN’s Landscape Assessment (also known as AusCover) facility have been collected via satellite remote sensing, typically using Landsat or MODIS (but also SPOT-5; see above). In some of the data products, satellite imagery has been combined with information obtained using other technologies, such as lasers or radars, to produce the resulting data product. Finally, a small number of TERN’s cover data products have been either exclusively or partially collected in the field. These data can be used for reference and validation of the remotely sensed data. Another of TERN’s facilities, TERN Ecosystem Surveillance (also known as AusPlots), collects large amounts of fractional cover data all over Australia that can also be used for validation of the remotely sensed fractional/ground cover data.

   The tables below (Tables 1 and 2) describe TERN’s cover data products. They include details on their cover type (i.e. fractional ground, land, and ‘other’ cover), spatial and temporal coverage and resolution, as well as on the files containing the data. Further details on specific data products can be obtained from their metadata.

Table 1. Cover data products available from TERN Landscape Assessment (AusCover) facility, including their type (i.e. fractional cover, ground cover, land cover, or ‘other’ cover data products; see above), their source, spatial coverage, and temporal coverage.

Table 2. Description of the temporal and spatial coverage and resolution of the datasets in the ‘Cover’ Data Products in Landscape Assessment (AusCover) facility, as well as the description of the files containing the datasets. In pink data products that currently can be downloaded. Acronyms: CRS = Coordinate Reference System, nc = NetCDF.

REFERENCES

• Australian Bureau of Agriculture and Resource Economics and Sciences (ABARES) and Food http://www.agriculture.gov.au/abares/aclump/pages/definitions.aspx
• Pete Fisher; et al. (2005). "Land use and Land cover: Contradiction or Complement". In Peter Fisher, David Unwin. Re-Presenting GIS. Chichester: Wiley. pp. 85–98.
• Food and Agriculture Organization of the United Nations (FAO) http://www.fao.org/docrep/003/x0596e/X0596e01.htm