Landcare Research - Manaaki Whenua

Landcare-Research -Manaaki Whenua

EcoSat and LCDB

Severe erosion in southern Hawkes Bay following a major storm in 2011. Image - Les Basher

Severe erosion in southern Hawkes Bay following a major storm in 2011. Image - Les Basher

What is the difference between EcoSat and LCDB (Land Cover Database)?

The table below contrasts EcoSat with the LCDB. Although both systems use satellite imagery to produce land information at 1:50 000 scale, they are different in many aspects. LCDB was designed primarily to report on changes to land cover and therefore has needed to maintain the mapping standards established for the LCDB1 (e.g., 1 ha minimum mapping unit). EcoSat in contrast has been free to use new technology and develop different standards for a variety of products. These products have been designed to complement LCDB. For example, EcoSat basic land cover maps actual land cover, whereas LCDB focusses more on land use (e.g., recently cleared forest in an exotic forest estate is mapped by EcoSat as bare ground, whereas it is mapped by LCDB as exotic forest). EcoSat does not attempt to produce a single snapshot in time, like LCDB, but has built up a national coverage of ETM+ imagery over 4 years. Thus it has been possible to archive high quality imagery with relatively high sun angles and low cloud cover. The high sun angles have permitted the flattening of imagery and the automatic processing to products with 15-m pixel minimum mapping unit.

Summary table

EcoSat LCDB2
  • 15 m ortho-rectified imagery
  • 15 m standardised spectral reflectance (calibrated and flattened imagery)
  • 15 m woody vegetation layer
  • 15 m basic land cover (15 classes)
  • 15 m ecosystem energy absorption
  • 15 m water courses and riparian vegetation
  • 15 m DEM
  • 15 m indigenous forest composition
  • 15 m wetlands (16 classes)
  • Corrected LCDB1 (18 classes)
  • LCDB2 layer (42 classes)
  • 15m ortho-rectified imagery
  • Land cover change map (1997-2002)
Minimum mapping unit
  • 225 m2
  • 10000 m2
Geo-registration accuracy  
  • 10 m
  • 10 m
Primary uses
  • Regional forest inventory
  • Possum control monitoring
  • Regional land-cover inventory
  • Regional riparian vegetation inventory
  • Land condition monitoring
  • Regional wetland inventory
  • National forest inventory
  • Kyoto protocol reporting
  • National land-cover inventory and 5 yearly change map/statistics
  • SOE reporting National–Local
  • National wetland inventory > 1 ha
  • Significant indigenous vegetation (RMA S6c)
Production methods
  • Atmospheric, directional reflectance, and topographically corrected imagery
  • Automatic per pixel classification
  • Objective reflectance rules
  • Manual drawing of LCDB1 polygons by visual interpretation
  • Corrected LCDB1 and LCDB2 change polygons automatically drawn or manually drawn
  • Intensive ground data and field verification required as not all classes are spectrally separable
Acquisition period
  • Best recent imagery over last 4 summers
  • Region by region approach
  • 2001–2002 summer imagery
  • National snapshot every 5 years
  • Includes off-shore islands
  • Regional partnerships contribute to cost of production. Ownership remains with LCR
  • A publicly accessible Crown database