No abstract

Sixteen years have passed since the last global volcanic event and more than 25 since a volcanic catastrophe that killed tens of thousands. In this time, volcanology has seen major advances in understanding, modelling and predicting volcanic hazards and, recently, an interest in techniques for reducing and mitigating volcanic risk. This paper provides a synthesis of literature relating to this last aspect, specifically the communication of volcanic risk, with a view to highlighting areas of future research into encouraging risk-reducing behaviour. Evidence suggests that the current 'multidisciplinary' approach within physical science needs a broader scope to include sociological knowledge and techniques. Key areas where this approach might be applied are: (1) the understanding of the incentives that make governments and communities act to reduce volcanic risk; (2) improving the communication of volcanic uncertainties in volcanic emergency management and long-term planning and development. To be successful, volcanic risk reduction programmes will need to be placed within the context of other other risk-related phenomena (e.g. other natural hazards, climate change) and aim to develop an all-risks reduction culture. We suggest that the greatest potential for achieving these two aims comes from deliberative inclusive processes and geographic information systems.

Mapping lava flows using satellite images is an important application of remote sensing in volcanology. Several volcanoes have been mapped through remote sensing using a wide range of data, from optical to thermal infrared and radar images, using techniques such as manual mapping, supervised/unsupervised classification, and elevation subtraction. So far, spectralbased mapping applications mainly focus on the use of traditional pixel-based classifiers, without much investigation into the added value of object-based approaches and into advantages of using machine learning algorithms. In this study, Nyamuragira, characterized by a series of more than 20 overlapping lava flows erupted over the last century, was used as a case study. The random forest classifier was tested to map lava flows based on pixels and objects. Image classification was conducted for the 20 individual flows and for 8 groups of flows of similar age using Landsat 8 imagery and a DEM of the volcano, both at 30-meter spatial resolution. Results show that object-based classification produces maps with continuous and homogeneous lava surfaces, in agreement with the physical characteristics of lava flows, while lava flows mapped through the pixel-based classification are heterogeneous and fragmented including much "salt and pepper 2 noise". In terms of accuracy, both pixel-based and object-based classification performs well but the former results in higher accuracies than the latter except for mapping lava flow age groups without using topographic features. It is concluded that despite spectral similarity, lava flows of contrasting age can be well discriminated and mapped by means of image classification. The classification approach demonstrated in this study only requires easily accessible image data and can be applied to other volcanoes as well if there is sufficient information to calibrate the mapping.

In this study, linear spectral mixture analysis (LSMA) is used to characterize the spectral heterogeneity of lava flows from Nyamuragira volcano, Democratic Republic of Congo, where vegetation and lava are the two main land covers. In order to estimate fractions of vegetation and lava through satellite remote sensing, we made use of 30 m resolution Landsat Enhanced Thematic Mapper Plus (ETM+) and Advanced Land Imager (ALI) imagery. 2 m Pleiades data was used for validation. From the results, we conclude that (1) LSMA is capable of characterizing volcanic fields and discriminating between different types of lava surfaces; (2) three lava endmembers can be identified as lava of old, intermediate and young age, corresponding to different stages in lichen growth and chemical weathering; (3) a strong relationship is observed between vegetation fraction and lava age, where vegetation at Nyamuragira starts to significantly colonize lava flows ~15 years after eruption and occupies over 50% of the lava surfaces ~40 years after eruption. Our study demonstrates the capability of spectral unmixing to characterize lava surfaces and vegetation colonization with time, which is particularly useful for poorly known volcanoes or those not accessible for physical or political reasons.

The age of past lava flows is crucial information for evaluating the hazards and risks posed by effusive volcanoes, but traditional dating methods are expensive and time‐consuming. This study proposes an alternative statistical dating method based on remote sensing observations of tropical volcanoes by exploiting the relationship between lava flow age and vegetation cover. First, the factors controlling vegetation density on lava flows, represented by the normalized difference vegetation index (NDVI), were investigated. These factors were then integrated into pixel‐based multi‐variable regression models of lava flow age to derive lava flow age maps. The method was tested at a pixel scale on three tropical African volcanoes with considerable recent effusive activity: Nyamuragira (Democratic Republic of Congo), Mt Cameroon (Cameroon) and Karthala (the Comoros). Due to different climatic and topographic conditions, the parameters of the spatial modeling are volcano‐specific. Validation suggests that the obtained statistical models are robust and can thus be applied for estimating the age of unmodified undated lava flow surfaces for these volcanoes. When the models are applied to fully vegetated lava flows, the results should be interpreted with caution due to the saturation of NDVI. In order to improve the accuracy of the models, when available, spatial data on temperature and precipitation should be included to directly represent climatic variation. Copyright © 2017 John Wiley & Sons, Ltd.