Working with the Ministry of Agriculture, Food, Forests, and Fisheries (MAFF) team in Vava'u, a workflow to map the Vanilla plantations on Vava'u was created. In May and June 2020 all the vanilla plantations on Vava'u were mapped to assess the area of the crop in well-managed or neglected state.
Read more about this work on the QField blog: Vanilla Surveys using QField
Using ESA Sentinel-2 images with a 10 m spatial resolution and 5 day revist period, a machine learning model has been trained to each 10 m pixel's land cover per year. Relative to other land cover maps with coverage over Pacific Island countries, this approach offers improved thematic resolution (it captures a broader range of land cover classes which reflects actual landscape heterogeneity - current maps used in Fiji with similar spatial resolutions typically map only forest and non-forest classes), improved spatial resolution (relative to other datasets with similar thematic resolution of land cover classes captured), and captures local-level inter-annual land cover changes (most land cover maps with Pacific Island coverage are either for a single snapshot in time or monitor land cover dynamics at coarser spatial resolutions).
Initial field validation has been undertaken in Ba, Fiji. Collaboration with the Ministry of Forests Resource and Assessment Group (FRAC) in Fiji is ongoing to support use of this land cover mapping method for REDD+ reporting requirements and using QField mobile GIS to collect ground truth data for model training and validation (training workshop on ground truth data collection).
Read more about this work on the QField blog: Ground Truth Data Collection Using QField for LULC Mapping in Fiji
Using QField, the MAFF team on Vava'u mapped the area of farmland utilised by each village's grower groups. These maps were used to allocate fuel to grower groups to help bring under-utilised land into cultivation given the backdrop of food supply chain vulnerability due to Covid-19. These maps were also presented to town officers (local community officials in Tonga).
In collaboration with the MAFF team from Vava’u, Tonga, mobile data collection forms and spatial databases have been developed for mapping the extent of crops grown within farms and monitoring farm management practices. This data is collected in-the-field using QField mobile GIS.
Using these tools, MAFF staff are mapping farms on the island groups of Tongatapu, Vava'u, Eua, and Ha’apai in Tonga. This field work is supported by web forms and server-side apps for data management and the map.landscape dashboard for data syncing and visualisation.
A household survey covering topics related to household liveihood activities, use of landscape resources, and climatic challenges to managing landscape resources was undertaken in four communities in Coastal Ba, Fiji - Etatoko, Koronubu Vunibaka, Natutu, and Nawaqarua.