[1] The distribution of frost table depths on a peat-covered permafrost slope was examined in a discontinuous permafrost region in northern Canada over 4 consecutive years at a variety of spatial scales, to elucidate the role of active layer development on runoff generation. Frost table depths were highly variable over relatively short distances (0.25-1 m), and the spatial variability was strongly correlated to soil moisture distribution, which was partly influenced by lateral flow converging to frost table depressions. On an interannual basis, thaw rates were temporally correlated to air temperature and the amount of precipitation input. Simple simulations show that lateral subsurface flow is governed by the frost table topography having spatially variable storage that has to be filled before water can spill over to generate flow downslope, in a similar manner that bedrock topography controls subsurface flow. However, unlike the bedrock surface, the frost table is variable with time and strongly influenced by the heat transfer involving water. Therefore, it is important to understand the feedback between thawing and subsurface water flow and to properly represent the feedback in hydrological models of permafrost regions.

Abstract:Hillslope runoff in permafrost regions covered by organic soil is strongly influenced by subsurface flow in the active layer, as well as surface flow where the active layer is very shallow. Flow rates in the organic-rich active layer are strongly dependent on the depth to the thawing front (i.e. frost table) and the corresponding soil hydraulic conductivity at that depth. Therefore, hydrological models for permafrost terrains need to simulate the thawing of the active layer accurately. In order to simulate the downward movement of the frost table, a simple heat-conduction model was proposed and compared to field data from a wet, organic-covered watershed in a discontinuous permafrost region of Canada. Ground heat flux was measured simultaneously using the calorimetric, gradient, and flux-plate methods to increase the confidence in data sets. The majority (>86%) of ground heat flux was used to melt the ice in frozen soil, and the soil temperature had a linear profile from the ground surface to the frost table when averaged over several days. Assuming a linear temperature profile, the proposed method calculates the daily rate of thawing from ground surface temperature and bulk thermal conductivity, where the latter is essentially determined by soil water content. Simulated depths to the frost table during three thaw seasons (2003)(2004)(2005) matched closely with the observed data for two contrasting ground-cover types with distinctly different thaw rates. The method can be easily implemented in hydrological models, and will provide a useful tool for simulating hillslope drainage in organic-covered permafrost terrains, and for evaluating the effects of topography and land cover on the temporal and seasonal variability of the frost table.

Abstract:Peat plateaus are important landscape features of many high-boreal, wetland-dominated drainage basins. Raised up to 2 m above the surrounding landscape and underlain by permafrost, these forested peatlands provide meltwater drainage to the surrounding wetlands, and to basin runoff. Understanding the factors that control the volume and timing of runoff from peat plateaus is the essential first step towards developing methods of accurately predicting basin runoff from wetland-dominated basins in the region of discontinuous permafrost, as well as understanding the basin response to hydrological changes brought on by the thermal degradation and thaw of permafrost peatlands. In this study, a water balance approach and the Dupuit-Forchheimer equation were used to quantify sub-surface runoff from a forested peat plateau at Scotty Creek, a small (152 km 2 ), wetlanddominated discontinuous permafrost basin in Northwest Territories, Canada. These two computations yielded similar results in both years of study (2004)(2005), and showed that runoff accounted for approximately half of the moisture loss from the peat plateau, most of which occurred in response to snowmelt inputs. The melt of ground ice was also a significant source of water during the study periods, which was largely detained in soil storage. Soil moisture conditions prior to soil freezing were a major factor controlling the volume of runoff from the hillslope. Sub-surface drainage rates declined dramatically after the snowmelt runoff period, when the majority of water inputs went to soil storage and evapotranspiration. The minimal lag between rain events and hydrograph response in both years suggests that much of the runoff produced from rain events is rapidly transported to the adjacent wetlands. These results give insight into how current climate warming predictions for northern latitudes could affect the hydrological response of forested peat plateaus, and the basins which they occupy.

The spatial distribution of benthic (up to 0.05 m depth) and hyporheic (0.25 and 0.5 m depth) macroinvertebrates from downwelling zones at the heads of riffles and upwelling zones at the tails of riffles was examined in two studies on a 4th order chalk stream in Dorset, England. In the first study, differences in benthic and hyporheic macroinvertebrate community composition between the head and tail of a single riffle were investigated. In the second study, a replicated design involving eight riffles was used to compare benthic and hyporheic macroinvertebrate community composition both between heads and tails of the same riffles and between riffles. In the first (single riffle) study there were significantly higher mean numbers of benthic invertebrates and families at the riffle head (715 individuals and 13.8 families per 0.0225 m 2 ) compared to the tail (192 individuals and 8.7 families). ANOSIM analysis also showed that the community structure of head and tail benthic samples was significantly different. In the second (replicated riffle) study, there were also significantly more benthic invertebrates at riffle heads (" x=594 per 0.0225 m 2 ) compared to tails (" x=417 per 0.0225 m 2 ), although this was not the case for families, and community structure also differed significantly between riffle heads and tails. In contrast, in the hyporheic zone, there were no significant differences between the total numbers of invertebrates in the riffle heads and tails, or between riffles, although a significant difference in family richness between riffle head and tail samples was identified in the first study. Community analysis revealed progressively poorer separation of riffle head and tail samples at 0.25 m and 0.5 m hyporheic depths. Whilst being able to identify clear differences in benthic communities from riffles heads and tails, the physically heterogeneous nature of the riffle habitats studied made it difficult to account for the consistent differences in macroinvertebrate communities observed with the physical variables measured.

Abstract. The University of Victoria Earth System Climate Model (UVic ESCM) of intermediate complexity has been a useful tool in recent assessments of long-term climate changes, including both paleo-climate modelling and uncertainty assessments of future warming. Since the last official release of the UVic ESCM 2.9 and the two official updates during the last decade, considerable model development has taken place among multiple research groups. The new version 2.10 of the University of Victoria Earth System Climate Model presented here will be part of the sixth phase of the Coupled Model Intercomparison Project (CMIP6). More precisely it will be used in the intercomparison of Earth system models of intermediate complexity (EMIC), such as the C4MIP, the Carbon Dioxide Removal and Zero Emissions Commitment model intercomparison projects (CDR-MIP and ZECMIP, respectively). It now brings together and combines multiple model developments and new components that have come about since the last official release of the model. The main additions to the base model are (i) an improved biogeochemistry module for the ocean, (ii) a vertically resolved soil model including dynamic hydrology and soil carbon processes, and (iii) a representation of permafrost carbon. To set the foundation of its use, we here describe the UVic ESCM 2.10 and evaluate results from transient historical simulations against observational data. We find that the UVic ESCM 2.10 is capable of reproducing changes in historical temperature and carbon fluxes well. The spatial distribution of many ocean tracers, including temperature, salinity, phosphate and nitrate, also agree well with observed tracer profiles. The good performance in the ocean tracers is connected to an improved representation of ocean physical properties. For the moment, the main biases that remain are a vegetation carbon density that is too high in the tropics, a higher than observed change in the ocean heat content (OHC) and an oxygen utilization in the Southern Ocean that is too low. All of these biases will be addressed in the next updates to the model.