On 21st and 22nd November 2016, natural disaster hit the French overseas territory of New Caledonia. Several days of exceptional rainfall caused deadly flooding and two devastating landslides in the Houailou district, 235km from the capital Noumea. In just one hour, the discharge of the Néaoua river rose from tens of cubic metres per second, to 600m3 per second, causing the worst flooding in living memory and scouring or silting up nearly 10% of the surface of the agricultural plain. Around a hundred landslides combined, forming mudslides in small catchment basins, engulfing ten homes, and leaving eight people missing.
Critical information over the affected area was urgently required in order to help first responders plan emergency response on the ground. Accurate and the freshest imagery possible was also immediately needed to further understand the impact of the event and assess the damage caused.
“Thanks to Airbus Defence and Space Pléiades constellation and our local partner INSIGHT, we could access rapid post-event imagery. In addition, the resolution and precision of the Pléiades data enabled us to fine-tune our analysis. The reactivity of the teams concerned, Airbus plus INSIGHT, meant that we obtained the information we needed in less than 48 hours!” Mr. Jean-Christophe Lefeuvre, Former Head of the Conservation International office in New Caledonia Géomatique.
SOLUTIONS AND RESULTS
Relying on Airbus’ satellite constellation’s high reactivity, the Pléiades satellites were tasked with the highest priority over the town of Houailou in order to collect two 50cm products of the affected area, with a minimum cloud cover and an acquisition angle compatible with the operational needs of the project. Promise kept, the images were delivered after the tasking activation.
Airbus Defence and Space reseller in the Pacific, INSIGHT, carried out automatic processing of the Pléiades data in order to extract and understand the changes which had taken place.
In order to qualify the field data more precisely the change detection solution, using Pléiades 50cm products, uses a number of information enhancement techniques:
– A co-registration phase that allows perfect superposition of the images used.
– A change detection phase, which involves three main steps:
1) The “Magnitude Difference” which is the conventional method for change detection, taking into account the time lapse between the two images and apparent changes in vegetation.
2) The Principal Component Analysis (PCA), which is used for more detailed characterization of the changes on the ground.
3) And the Subtractive method that enables further details to be determined. This method is, however, heavily influenced by shadows on the vegetation.
– Finally, the results were concatenated and vectorised for direct integration into the GIS of the Conservation International NGO.
All of these processing operations and quality control checks were made available to Conservation International by INSIGHT just 24 hours after Airbus Defence and Space delivered the source data.
Pascal Sawa, Mayor of Houailou said: “Thanks to the partnership with Conservation international, an independent NGO, and INSIGHT, a Caledonian service provider, the Town Hall received a certain amount of information and summary documents very rapidly. These documents analysed all the various data concerning the sector, providing extremely valuable information in terms of observation of the sequence of events and provided an accurate and reliable description of the phenomena involved. They offered numerous options for investigation directly linked to the search for those who were unaccounted for.”
- Clearer understanding of the impact of the rainfall on the ground and precise insights over the area affected by the disaster, due to accurate and reliable change detection based on Pléiades 50cm products.
- Rapid access to critical information thanks to Pléiades unrivalled responsiveness combined with INSIGHT’s reactivity.
- Precise information thanks to the excellent geometrical accuracy of Pléiades imagery and clearer assessment of the variability of the vegetation cover thanks to high quality Near-Infrared (NIR) products.