Work Package 4: Spatial Analysis and Modelling
Objective: spatial analysis and modelling approaches are used to investigate the controlling processes for the observed changes found in WP 3 for (A) Lava flows (B) Topography and hydrology (C) Glacial erosion (D) Plant community distribution and (E) SOC variability. In addition this WP ensures that these models are tuned and applied to Icelandic scenarios.
A) Lava flow modelling
Lava flow simulation is a technique used for hazard assessment and is of relevance for both inhabitants as well as tourists in the Hekla area. The pre-emplacement topography constitutes the computational domain and substantially influences the lava flow simulation output (e.g., Miyamoto and Papp, 2004; Tarquini and Favalli 2010) and the lava volume is another critical parameter (Walker 1973, Malin 1980). These parameters are derived for four Hekla eruptions (WP. 3.2).
Along with observational data they provide a unique possibility to calibrate the lava flow models to Hekla and correlate lava flow morphology with lava flow modelling. This WP relies on the expertise in lava flow modelling developed over several years at INGV. Given a reliable pre-emplacement topography and the planimetric extent of historical lava flows, it is possible to calibrate the DOWNFLOW code (e.g.,Tarquini and Favalli, 2011). Hence, the path of future eruptions can be derived by considering a probability density function for future vent opening (e.g., Tarquini and Favalli 2013) and the high resolution topography derived from the LIDAR data. In addition to the above, the new code F-L is already in its final refinement phase, and by using this new-generation probabilistic code it is possible to account not only for the planimetric extent of lava flows, but also for the style of emplacement (Glaze and Baloga, 2013; Hamilton et al. 2013)
B) Terrain and Hydrological modelling (MadAlgo, IES.EcoInf: G.B.M. Pedersen)
Terrain and hydrological modelling provide a tool to investigate the topography and hydrology as potential driving mechanisms for observed landscape changes. This task of the project will be carried out at MadAlgo in cooperation with EcoInf, who previously have carried out similar analysis (e.g., Moeslund et al., 2013):
1) Calculation of mechanistic topographic factors such as: terrain slope, terrain curvature, terrain aspect, terrain roughness and potential solar radiation using Spatial Analyst tool in GIS. Potential wind exposure (Boose et al. 1994, Mikita and Klimánek 2010) will be calculated by using spatial analyst hillshade tool together with meteorological data on wind speed and direction.
2) Hydrological modelling by using SCALGO Hydrology software (SCALGO, Aarhus, Denmark) as well as using GIS to model (1) flow direction (2) flow accumulation and drainage, (3) Catchment areas and (4) topographic wetness index (Wilson and Gallant, 2000).
C) Glacial erosion modelling (GS, IES: D.L. Egholm & G.B.M. Pedersen)
Öræfajökull provides a unique setting to explore the glacial erosion processes that drive feedbacks between topography and glacial dynamics. By rapid abrasion and quarrying of the underlying bedrock, glaciers are known to exert a first-order control on the elevation of Earth’s mountain ranges. Glacial flow dynamics and subglacial erosion will be modeled on a 20-30m grid scale under conditions of rapid uplift and short time-scales of glacial isostatic rebound. This grid-scale allows us to link the landscape changes measured in WP 3.2 to erosion processes using state-of-the-art computational landscape evolution
models. Based on the Öræfajökull DEM, a computational higher order ice sheet model coupling flow equations for ice, water, and sediment with glacial, periglacial, and fluvial erosion (Egholm et al., 2012) will be used to simulate long-term landscape evolution. The unique data sets provided by EMMIRS will allow us to test the model assumptions used and to identify the most important processes and feedbacks. The simulations will have focus on the evolution of overdeepenings and on links between subglacial bedrock erosion and periglacial hillslope processes. This part of the project will be carried out at GS, AU, Denmark.
D) Modelling vegetation
Spatial distribution of vegetation is determined by a number of environmental factors such as elevation, moisture, eolian activity, topography and the underlying geology (e.g. Houle, 1997; Jumpponen et al., 1999; Moeslund et al., 2013) and function on different scales. By using the vegetation map and the LIDAR DEM, the effects of the various environmental factors on vegetation distribution can be analyzed in order to estimate the relative influence of the different factors on the spatial distribution of plant communities. The factors investigated could include elevation, soil moisture, slope gradient, slope aspect, slope shape, wind exposure and incident solar radiation (derived in WP 4.3B). This part of the project will be carried out at EcoInf, AU, Denmark.
E) Modelling SOC (LES, IINH, IFR, AUI, SCSI: G. Gísladóttir & O.K. Vilmundardóttir).
Similar to the spatial distribution of plant communities, the underlying soils are affected by landscape where soil moisture and erosional/depositional processes impact the evolving soil properties (e.g., Egli et al., 2006; Yoo et al., 2007; Gísladóttir et al., 2010, 2011; Vilmundardóttir et al., 2014, in review), causing high spatial variability of SOC. For spatial modelling of the SOC stock, a soil sampling strategy will be structured with regards to vegetation, geomorphic and topographic features. Sampling requires the use of the (i) vegetation map, (ii) the geologic map including geomorphologic features, and (iii) the topography, e.g. slope gradient, slope aspect and slope curvature to estimate the mean SOC stock under the plant communities and surface types relevant to the study areas with an acceptable degree of certitude. From this a regional assessment of the SOC stock can be done and presented visually on a map, a process not previously attempted in Iceland. Soil samples will be analyzed at the University of Iceland (UI) for SOC content and bulk density for calculations of SOC stock. By combining the regional carbon stock assessment of the aboveground vegetation (derived in WP 2.3B) and the SOC stock assessment the regional carbon stock of the terrestrial ecosystem (vegetation and soils) can be evaluated. Additionally, modelling rates of SOC accretion is possible through sites of age-chronosequences on lavas around Hekla and along the recessional paths of the Öræfajökull’s outlet glaciers where the location of termini is known in time as done by Vilmundardóttir et al. (2014).