Secondary Forest Detection in a Neotropical Dry Forest Landscape Using Landsat 7 ETM+ and IKONOS Imagery1

Authors

  • Juan Pablo Arroyo-Mora,

    Corresponding author
    1. Department of Earth and Atmospheric Sciences, Earth Observation Systems Laboratory, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • G. Arturo Sánchez-Azofeifa,

    Corresponding author
    1. Department of Earth and Atmospheric Sciences, Earth Observation Systems Laboratory, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • Margaret E.R. Kalacska,

    1. Department of Earth and Atmospheric Sciences, Earth Observation Systems Laboratory, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • Benoit Rivard,

    1. Department of Earth and Atmospheric Sciences, Earth Observation Systems Laboratory, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • Julio C. Calvo-Alvarado,

    1. Instituto Tecnologico de Costa Rica, Department of Forest Engineering, Cartago, P.O. Box 159-7050, Costa Rica
    Search for more papers by this author
  • Daniel H. Janzen

    1. Department of Biology, University of Pennsylvania, Philadelphia, PA 19104, U.S.A.
    Search for more papers by this author

  • 1

    Received 15 October 2004; revision accepted 8 May 2005.

2   Current address: Department of Ecology and Evolutionary Biology, 75 North Eagleville Rd, Unit 3043, University of Connecticut, Storrs, CT 06269, U.S.A.

3   Corresponding author; e-mail: arturo.sanchez@ualberta.ca

ABSTRACT

We integrate forest structure and remotely sensed data for four successional stages (pasture, early, intermediate, and late) of a tropical dry forest area located in the Sector Santa Rosa of the Guanacaste Conservation Area in northwestern Costa Rica. We used a combination of spectral vegetation indices derived from Landsat 7 ETM+ medium resolution and IKONOS high-resolution imagery. The indices (using the red and near-infrared bands) simple ratio and normalized difference vegetation index separated the successional stages well. Two other indices using mid-infrared bands did not separate successional stages as well. In a comparison of the successional stages with chronological age, there was no separability in the spectral reflectance among different age classes. Successional stages, in contrast, showed distinct groups with minimal overlap. We also applied a simple validation in another dry forest located in the Palo Verde National Park in the province of Guanacaste, Costa Rica, with reasonably good results.

RESUMEN

En este estudio, datos de campo y datos remotamente sensados son analizados para cuatro etapas sucesionales (pastos, temprano, intermedio y tardío) de un bosque seco tropical ubicado en el sector Santa Rosa del Area de Conservación Guanacaste en el noroeste de Costa Rica. Utilizamos una combinación de indices spectrales derivados imágenes satelitales, de mediana resolución de Landsat 7 ETM+ y de alta resolución de IKONOS. Los índices (usando las bandas de rojo e infrarojo cercano) de relación simple (SR) e índice normalizado de diferencia vegetal (NDVI) separó bien las etapas sucesionales. Así otros dos índices, usando las bandas de infrarojo medio no separaron las etapas sucesionales. En una comparación de las etapas sucesionales con la edad cronológica, no existió separabilidad en la reflectancia espectral entre las diferentes clases de edad. Sin embargo, las etapas sucesionales mostraron grupos distinguibles con mínima sobreposición.También aplicamos una simple validación en otro bosque seco ubicado en el Parque Nacional Palo Verde en la provincia de Guanacaste, Costa Rica.

Ancillary