Northeastern US Landcover Classification from AVHRR

Eleven landcover categories classify the Northeastern US region, including urban, agriculture, and multiple forest types. Scott Ollinger created this 1-kilometer resolution dataset by combining classified AVHRR imagery with USGS Land Use data. The data supported a regional forest change modeling effort led by Lathrop and Bognar.

Use Cases

• Classify pixels into 11 landcover categories, such as 'hardwoods' or 'mixed hardwood/spruce fir', for regional land use mapping.
• Estimate total forest area using the 'hardwoods', 'mixed hardwood/spruce fir', and 'mixed hardwood/pine' classes for comparison with county-level statistics.
• Enhance discrimination between coniferous and deciduous forest types using AVHRR imagery recorded after leaf fall.
• Combine the classified landcover data with other GIS layers, such as the USGS Land Use/Land Cover data used in its creation, for integrated environmental modeling.

Strengths

• Classification distinguishes 11 specific landcover categories, including multiple forest types.
• Statistical analysis showed forest area estimates were comparable to U.S. Forest Service county-level data.
• AVHRR imagery from post-leaf-fall periods enhanced coniferous vs. deciduous forest discrimination.

Limitations

• Spatial resolution is approximately 1 kilometer, limiting analysis of fine-scale landcover patterns.
• Geographic scope is restricted to the Northeastern US (eastern New York and New England).
• Dataset age and update frequency are unknown, potentially limiting relevance for current conditions.

Provenance

Source

NASA Earthdata via SCIOPS.

Collection Method

Created by combining single-date classified AVHRR imagery with existing USGS Land Use/Land Cover data.

Geography

Northeastern United States, encompassing eastern New York state and all of New England.