GfGD Conference – Selected Poster Abstracts (Part 1)

An important part of our annual conference is giving students and recent graduates the opportunity to present their work – through a poster and drinks reception at the end of the day. Last year we had some fantastic posters – from both undergraduate and postgraduate students, and recent graduates. This week and next we’ll be publishing the titles and abstracts for some of the posters being displayed at this years conference.

**Register now for your ticket**

—–

Investigating controls on river morphology across the Ganga Plains, northern India.

Elizabeth Dingle, Hugh Sinclair, Mikael Attal (University of Edinburgh)

Many of the rivers of the Ganga Plains are prone to abrupt switching of channel courses (avulsion) causing devastating floods over some of the most densely populated regions on the globe. During 2008, a single channel avulsion event resulted in a 120 km eastward shift of the River Kosi resulting in extensive flooding and devastation to the millions of people living in these areas. Despite this, our understanding of the factors that control the dynamics of these river systems is surprisingly limited.

We have developed a novel approach to quantifying floodplain and channel relief that easily identifies areas where channels are perched relative to their adjacent floodplain. This is important as areas lower than the channel are most at risk of inundation should the channel avulse. Here we present the results when this tool is applied across major rivers draining the Himalayan foreland basin. Overall the degree of channel entrenchment, and therefore channel stability, increases from east to west across the Ganga Plains. The nature of incision across the Ganga Plains is highly variable however, even between adjacent systems suggesting contrasting or multiple controls on river morphology. This includes differences in the downstream evolution of incised valley depth and width. The next challenge is to isolate potential controls behind the observed river morphology, which will be aided with extensive sediment grain size and sediment flux data being collected later this year. These results highlight the necessity of considering the morphology of these rivers as a function of a coupled mountain-basin system. This is fundamental to determining the sensitivity of river dynamics to projected climate change and anthropogenic development on the plains.

—-

Volcanoes, coconuts and flying foxes: A month of work experience in Vanuatu

Ben Clarke and Eleri Simpson (University of Leicester)

Vanuatu is a volcanically active archipelago about 1000 miles north of New Zealand. We  arranged to travel there and spend one month working with the Vanuatu Geohazards Observatory. The people at the observatory are kind and hard working, but with little funding it is difficult for them to fully observe all of the 6 active volcanoes. Our job was to venture to Lopevi volcano, a potentially hazardous stratovolcano in the centre of the archipelago. Currently, Lopevi is only monitored by one seismometer which does not relay the data in real-time; so very little is actually known about its current activity. We created a map of the newest lava flows and other deposits, meanwhile teaching their volcanic data analyst everything we knew about mapping. In addition to this, we spoke to the locals and found out a lot about the social issues surrounding the volcano, including the potential return of former inhabitants, displaced by a large eruption in 1963. When we returned to the capital city (Port Vila) we spent the remaining time digitising the map and creating public information brochures informing the general population about what to do in the event of a Tsunami, Earthquake or Volcanic eruption. Although there were many challenges such as cultural, dietary and language differences, it was a highly worthwhile experience for both us and the observatory. 

—-

Modelling landslide susceptibility in a data sparse region: Papua New Guinea 

Joanne Robbins (University of Leicester)

There has been a substantial volume of research developing and validating approaches for landslide susceptibility, most of which utilise geo-information and geographic information system (GIS) tools. Methods to understand landslide susceptibility generally follow a two-step process whereby: (1) landslides are identified and classified within a historical landslide inventory and (2) environmental causal factors are identified and classified before being integrated to produce landslide susceptibility maps. Frequently this approach requires a wide range and number of datasets in order for the susceptibility maps to be representative and robust. In Papua New Guinea (PNG) landslide recording is limited and datasets applicable to landslide susceptibility analysis are sparse. Therefore in this research, satellite (Landsat Imagery) and airborne (GeoSAR (synthetic aperture radar)) technologies have been used to produce new, detailed landslide inventory maps for two case study regions in PNG (Chimbu and Western Province). Furthermore, these data have been used to undertake assessments to identify the principal control factors influential in causing slope movements in the two case study areas.

191 landslides were mapped for the Western Province case study domain and 366 landslides were mapped in the Chimbu Province domain. The location of each landslide head scarp was assessed against the spatial distribution of lithological (lineaments, drainage, rock type and land cover) and topographical (elevation, slope, aspect and curvature) characteristics and the dominant environmental control factors for each case study area were identified through frequency ratio analysis. Based on these statistics, fuzzy membership maps were produced. Different sets of maps, representing different environmental control factors, were combined through fuzzy relation-based approaches, to produce a range of susceptibility maps with differing degrees of accuracy when compared against the observed locations of landslides. The best maps for each case study region were identified through 2×2 contingency table analysis.