Defining Surface Land Cover Features Using High Resolution Imagery from Unmanned Aerial Systems

The ability to quickly collect high resolution imagery using Unmanned Aerial Systems (UAS) is known to be useful for a variety of survey and monitoring applications; however, the potential applications in earth and atmospheric science research have yet to be fully explored. This potential is considerable given the high spatial resolution available from the resultant imagery and the wide range of instrumentation that could be used on such platforms. To examine the research potential of data derived from UAS imagery, this project utilizes imagery from a series of UAS missions flown over southern Louisiana to develop a method for identification and classification of land cover and surface water features. Currently, the system is being flown over the Pearl River in Slidell, LA every two months. This provides a seasonal record that is important for identifying both the temporal changes in vegetation and the changes to the stream network. This information is vital for subsequent research on surface hydrologic processes and land surface/atmosphere interactions, with the specific advantages of using UAS evident through higher precision stream identification and land cover classifications. Oftentimes, accurate identification of streams and surface vegetation in coastal watersheds is difficult due to the dynamic nature of the environment, especially after extreme hydrological or meteorological events (i.e., tropical storms, river flooding, storm surge, etc.); therefore, the flexible temporal resolution of UAS is another definite advantage. Results of this project will include a method for extracting land surface features and characteristics, which will be used for future research in land surface/atmosphere simulations, surface water quality/quantity analysis, and vegetation dynamics.