New Zealand Hydrological Society (NZHS) Conference: A first dip into identifying small-scale wetlands using orthophotos, satellite imagery and lidar


The New Zealand Hydrological Society (NZHS) aims to promoting hydrological research, promoting education in all aspects of hydrology, disseminating information through the publication of a scientific journal, newsletters, and other appropriate means. The 2015 NZHS annual conference was themed ‘From Data to Knowledge’ and the society’s mission is to further the science of hydrology and its application to the understanding and management of New Zealand's water resources. I had the chance to present my latest work in the conference track “Monitoring and Data Management” on:

“Identifying small-scale wetlands using orthophotos, satellite imagery and LiDAR”

Wetlands play a key role in controlling flooding and non-point source pollution. In New Zealand, it has been estimated that over 90% of the former wetland area has been lost within a century and a half and there is an ongoing trend on wetland ecosystems. This increases the need for detailed identification and assessment of wetlands which would also improve our understanding of their ecological functioning. However, the inventory and characterization of wetland habitats are most often limited to small areas or done in a coarse scale. The diversity of wetlands makes it challenging for both the field scientists and the remote sensing analysts to identify and inventory these. The aim of this study was to evaluate the combination of multispectral imagery (Landsat), orthophotos and LiDAR data to precisely map the distribution of small-scale wetland habitats.
Study area in Southland, New Zealand
 

Fusion of high resolution orthophotos (0.4 m), Landsat imagery (30 m) and LiDAR based terrain indices (Topographic Wetness Index) were used to identify small-scale wetlands in the Waituna catchment in the Southland region. In addition, field observations were collected at the study site. Two radiometrically and geometrically corrected Landsat 8 images were acquired from December 2013 and August 2014 for multispectral analysis. High resolution orthophotos and topographic indices were used to increase the level of detail. Unsupervised and supervised classification was performed in ENVI 5 and spatial analysis in ArcGIS 10.3.
 

Topographic Wetness Index calculated for identifying wetlands

 
LANDSAT images alone didn’t give good results because of their coarse resolution. Wetlands often also have very complex boundaries and the fusion of multispectral imagery with high resolution orthophotos enabled to delineate wetland boundaries in more detail. In addition, LiDAR provided highly detailed information on the topography. In wetland ecosystems, the soil moisture is strongly correlated with the topography and consequently with vegetation distribution. Classification accuracy was highly improved when combining LiDAR data and multispectral images, enabling to delineate wetlands with the size of 0.1-02 ha.

I would kindly like to acknowledge my co-authors Chris Tanner and Sandy Elliott from NIWA and Ülo Mander from the Department of Geography, University of Tartu, Tartu, Estonia and Hydrosystems and Bioprocesses Research Unit, National Research Institute of Science and Technology for Environment and Agriculture (Irstea), Antony, France.
Location: Uni Waikato, Hamilton, New Zealand

Popular posts from this blog

GIS day

Image
GIS Day is playing a powerful role in creating geographic awareness everywhere in the world.  More than 80 countries are participating in holding local events such as corporate open houses, hands-on workshops, community expos, school assemblies, and more. Since 1999 GIS Day has been celebrated in Estonia. There are many interesting events organized on the day: conference, teachers computer lab with GIS courses, exhibition of companies and GIS related institutions, map contest, sightseeings to exhibitions by National Library of Estonia etc. This year I made presentation on the conference on topic "Mapping and modelling wetlands with GIS"   GIS Day is a good opportunity to show the usability and usefulness of GIS science to the public and especially to high school students and teachers. Presenting at GIS Day in National Library. Photo by Aivo Vard
Read more

Looking for wetlands on Routeburn Track

Image
Towards the  end of my stay in New Zealand I took the opportunity to walk one of the Great Walks, the Routeburn Track, which is location in the Mount Aspiring and Fiordland National Parks, south-west of the South Island. For the whole track over the distance of 32 km which I walked in three days. The track winds through beautiful beech forest, across tussock-covered flats, over the Harris Saddle, passes subalpine shrublands, alpine tarns and bogs. Most interestingly, even here science follows on the footsteps. I met a couple from Thames, Coromandel. They were dairy farmers and these few days on the track were their only only holidays for the last year. We had a great conversation about sustainable agriculture and manure management. They had apparently even won an industry award for best effluent management and  they collaborated with scientists and an innovative company that developed a solution to reduce the smell of a manure. That was a great example of scientists, entrepreneur
Read more