Analysis of Landsat 8 detection of the interannual variability of grassland curing in Greater Melbourne, Australia

Abstract

As one of the main components of Grassland Fire Danger Index, grassland curing degree provides crucial information for determining grassland fire danger. Accurate estimates of grassland curing are critical for determining grassland fire risk. This research focuses on the use of Landsat 8 to estimate grassland curing. Results demonstrate that Landsat 8 observations can be used to estimate curing percentages as assessed by visual and ground sampling measurements. Grassland interannual variability for the Greater Melbourne region using Landsat 8 imagery from 2013 to 2019 is examined. Slight differences in curing times and degree are observed for sample sites surrounding Greater Melbourne due to climatic differences across the region. Precipitation is regarded as an essential variable affecting curing degree and this relationship is evident for all five sample sites. Landsat 8 curing results are compared to both visual observations and destructive sampling, the most accurate method, for accuracy assessment. At 95% confidence level, Landsat 8 estimations are no different from destructive ground sampling estimations. Overall, this study validates the use of Landsat 8 data as an effective and accurate way for grassland curing monitoring.

Keywords

• grassland curing
• remote sensing
• grassland fires
• melbourne
• interannual variability