Reaney, S. M., Lane S. N. and Heathwaite A. L. 2009: Simulating the Spatial Distribution of Hydrological Connectivity Under Possible Future Climates – Impacts on River Flow Dynamics and Non-Point Source Pollution; American Geophysical Union Fall Meeting, San Francisco, USA
Predicted changes to the climate have the potential to alter the hydrological dynamics of catchments in response to new rainfall and temperate patterns. Primary effect will be on soil moisture and river flow. However, an important secondary effect will be on hydrological connectivity within the landscape. This factor is important since the locations of critical source zones for non-point source pollution are determined by the pollutant source risk pattern and the structure of the surface hydrological connectivity. The surface hydrological connectivity is determined by the topography and the temporal structure of the rainfall, which is predicted to alter under climatic change.
Using spatial connectivity statistics and a hydrological model, the potential for changes has been investigated. The hydrological simulations were performed using CRUM3, a fully distributed, physically based, model with a grid spatial structure (Reaney et al. 2007, Lane et al. 2009). The simulated catchment hydrological connectivity was based on the Network Index (Lane et al. 2004), which describes the propensity for each point in the landscape to connect to the receiving waters. From the Network Index, two connectivity indices were calculated: 1 the percentage of time that a point is connected to the receiving waters; and 2 the average number of connection – disconnection cycles per year. The percentage of time that a point is connected for give information of the potential for the export of non-source limited pollutants, such as fine sediment. The number of connection-disconnection cycles is important for pollutants that require aerobic conditions to accumulate and saturated conditions to be exported, such as nitrate.
Simulations were performed for a baseline period (1960 – 1990) and a climate change period (2071 – 2090) based on the medium-high emissions scenario from UKCIP02. The climate predictions were downscaled using the EARWIG weather generator (Kilsby et al. 2007). To gain insight into the uncertainty associated with the predictions, four realisations from the stochastic weather generator and four model parameter sets were used giving 16 simulations of 30 years per time period.
The simulation results show that 1, there may be significant changes in the size of the areas that will readily connect to receiving waters; 2, the duration of connection will increase; and 3, the number of connection – disconnection cycles will increase. These changes have the potential both the hydrological dynamics of the catchment and the sources and export of non-point source pollutants. Due to the non-linear relationship between catchment runoff and connectivity, there is the potential for increased high flows. The predicted changes to the surface flow connectivity have the potential to alter the location of critical sources areas for non-point source pollution and the distribution of pollution of non-point sources across the year.
Kilsby et al . 2007: Environ. Modell. Softw., 22, 1705-1719.
Lane et al. 2004: Hydrological Processes 18 (1): 191-201
Lane et al. 2009:Water Resour. Res., 45, W08423
Reaney et al. 2006: Hydrological Processes 21 (7), 894 – 906
Presented at the AGU Fall Meeting 2009 in San Francisco