Land monitoring on Mistake Creek Station
: evaluating the use of historically collected ground data to validate multi-temporal satellite data

  • Birgit Hallenstein

    Student thesis: Coursework Masters - CDU

    Abstract

    Pastoralism is a major land use in the Northern Territory. Monitoring vegetation change is important for managing rangelands and ensuring environmentally sustainable land use practices. Due to the vast extent of the Northern Territory pastoral estate, land managers are increasingly relying on remote sensing for monitoring cover change. Ground truthing plays an important role in understanding ecological processes and interpreting remotely sensed data.

    This study sought to evaluate whether historically collected ground monitoring data could be used to validate remotely sensed data, allowing for extrapolation of land condition interpretations up to the present day. The ground data used were collected from 1978-1993 to monitor vegetation change in response to rehabilitation works, following extensive degradation due to overstocking, on Mistake Creek Station in the Victoria River District.

    Satellite data were derived from a multi-temporal dataset of Landsat (MSS and I'M) images, by extracting cover index values in the visible red band, enabling a graphical representation of cover change over time. These results were compared to land condition classes determined from the ground data, and then integrated with climate history and management records to help explain land condition at the different monitoring sites.

    The results revealed that the satellite data did generally reflect the condition classes derived from ground data, responding in predictable patterns to favourable climatic conditions or disturbance events such as lire and grazing. Where remotely sensed cover indices did not respond as expected, explanations could commonly be attributed to changes in management practices or site specific differences, such as variation in soil type. These findings confirm that integrated monitoring systems can provide useful information to land managers using a variety of available data.
    Date of AwardNov 2004
    Original languageEnglish
    SupervisorDiane Pearson (Supervisor)

    Cite this

    '