Tools for soil erosion mapping and hazard assessment: application to New Caledonia, SW Pacific

Soil erosion is a major issue around the world. In the Pacific Islands, erosion has a strong impact on terrestrial and coastal ecosystems (mountains, alluvial plains, mangroves, coral reefs). Bush fire, deforestation and/or human activities accelerate erosion in mountainous areas especially on lateritic soils. Frequent and intense precipitation events (tropical depressions, cyclones) regularly strip particles from these sensitive soils. Any eroded sediment is quickly transported to coastal plains and to the sea along the main drainage lines, with immediate and recurrent impact on human activities such as open-cast mining, farming and fishing. There is therefore a need to identify key components of erosion processes for sustainable environmental management and planning. Particularly, we need to identify where the main erosion areas are, and what areas are impacted by severe sedimentation. If these areas are properly identified, management and mitigation can become effective. Hazard assessment is also required to plan new developments with limited erosion impacts. However, both mapping of active erosion and the assessment of hazard at a regional scale is time-consuming, costly and rarely updated. In that respect, two approaches have been tested in New- Caledonia:  Fast mapping of active erosion-linked areas: a Geographical Information System (GIS) based method was developed to model expert analysis from satellite images and topographical data. Bare ground surfaces were first mapped by remote sensing. Automated expert rules were then used to label pixels into different classes in relation to erosion and sedimentation. This method is fast, using only a few thematic layers. It provides a baseline for mapping inventory and for future updates that are essential for monitoring and hazard modelling. The quality of the mapping results are reasonable at the regional scale: the map shows zones of sediment production or accumulation, and aggravated erosion zones such as trails or open pit mines. The few identified errors seem to be easy to correct.  Optimisation of hazard modelling: a multidisciplinary approach based on data mining and geological knowledge was developed to efficiently assess erosion hazard. The method uses a readily available dataset used in an expert study area and a remote sensing product as training data. The results were consistent with general and local knowledge of erosion, suggesting that data mining is relevant for erosion hazard study. Furthermore, predictive models of erosion (i.e. hazard assessment) built with data mining methods seemed to give more realistic results than expert assessment. These preliminary results suggest that erosion assessment could become more efficient in the near future in New Caledonia. The mapping method is rapid and efficient and identifies erosion-linked areas in the most sensitive geological regions. Improvements to the method have been identified and a more detailed typology will make classification even more attractive. Likewise, data mining analysis could also be a useful tool for hazard assessment at the regional scale, even with sparse data and a partial inventory. Processing time for basic mapping was greatly reduced and hazard assessment improved with these methods. These semi- automated approaches will lead to better general and local knowledge.