05 Nov

Subpixel land cover mapping for an arid city

Posted in: Geospatial Analysis

Quantify urban expansion (1990-2011) of an arid city using sub-pixel land cover information of Landsat TM satellite imagery and Random Forest machine learning algorithm.


Land use land cover change; Sub-pixel modeling; Machine learning; Random forests algorithm; Dark object subtraction; vegetation mapping; Mapping vegetation dynamics of an arid city

Study area

Zhongwei, Nortwestern China Mapping vegetation dynamics of an arid city: study area

Research strategies

  1. Conduct sub-pixel vegetation mapping (1990-2011) with Random Forest (RF) machine learning algorithm by integrating high (OrbView-3) and medium spatial resolution (Landsat TM) data;
  2. Examine Dark Object Subtraction (DOS) atmospheric correction method to support spatial / temporal generalization of sub-pixel mapping algorithm;
  3. And characterize patterns of vegetation cover dynamics based on change detection analysis.

Mapping vegetation dynamics of an arid city: workflow

(Update: November, 2014)