Visual image interpretation of land use - a GI research perspective Dr Nigel Trodd Coventry University.

Visual image interpretation of land use - a GI research perspective

Dr Nigel TroddCoventry University

Land use human employment of the land

‘Almost all of the world’s lands are now used and managed’ (Richards 1990).

‘Human actions rather than natural forces are the source of most contemporary change in the states and flows of the biosphere. Understanding these actions and the social forces that drive them is thus of crucial importance for understanding, modeling and predicting global environmental change and for managing and responding to such change’ (Turner & Meyer, 1994).

Land use - data sources

‘Land use maps have been completed for many specific areas of the world. These efforts were based on various data sources and are not generally compatible with each other. A consistent, hierarchical classification system is ... to be encouraged’ (WMO 1997).

An ‘objective system that can be updated annually may be possible using remote-sensing of land cover change (perhaps combined with agriculture and forestry production statistics, socio-economic spatial data such as population and transportation). Landsat (TM) is one of the most appropriate data streams for land use monitoring. While it is still a research topic on how to use this data to produce a land use product, it is important that this data stream continues to be archived and kept accessible’ (WMO 1997).

Land use - research agenda

What information is needed?Frequency of measurement: Once every 5 years.

Spatial resolution: 5 m - 1 km depending on the spatial heterogeneity of land use; at least 30 m for many regions.

Accuracy/precision required: TBD.

What R T D is needed? Develop regionally-specific relationships between land cover

and land use;

Define the lowest acceptable spatial resolution;

Develop reliable procedures of inferring land use from land cover, on a regional basis.

Land use / land cover physical state of the land

single land use type may correspond to a single land cover classe.g. pastoralism to unimproved grassland

single land cover class may support multiple usese.g. forest used for timber, hunting/gathering,

fuelwood, recreation, wildlife preservation, soil protection

single land use may occupy multiple classese.g. mixed farming uses cultivated land, improved

pasture, settlements and woodland

Visual image interpretation

tone

texture

shadow

associati

on

shapepattern

size

site

Information products

Feature extraction

Information products

Integrated terrain unit

Information products

Land systems

Anderson - Level 1

1 Urban or built-up land

2 Agricultural land

3 Rangeland

4 Forest land

5 Water

6 Wetland

7 Barren land

8 Tundra

9 Perennial snow or ice

Anderson - Level 2

1 Urban or built-up land

11 Residential

12 Commercial & services

13 Industrial

14 Transportation, comms & utilities

15 Industrial & commercial complexes

16 Mixed urban or built-up land

17 Other urban or built-up land

Anderson - Level 2

2 Agricultural land

21 Cropland & pasture

22 Orchards, vineyards etc.

23 Confined feeding operations

24 Other agricultural land

Mapping land use

identify target land use categories define the properties of each land use

category in terms of land cover characteristics specify technique for extracting land cover

characteristics using elements of visual image interpretation

apply technique invert relationship between land cover and

land use generate land use information