Skip to main content

Integrating isotopic techniques with Bayesian modelling for improved assessment and management of global sedimentation problems

Periodic Reporting for period 1 - IMIXSED (Integrating isotopic techniques with Bayesian modelling for improved assessment and management of global sedimentation problems)

Reporting period: 2015-07-01 to 2017-06-30

Siltation presents a credible threat to river basin ecosystem service provision and water security. River silt originates, however, on catchment hillslopes and the primary driver for mobilisation and translocation downstream is soil erosion on agricultural land where loss of this finite resource threatens food security. Knowledge of sediment source and transfer dynamics in river catchments is critical to inform management policy decisions to maintain and enhance future food and water security. Identification of the source and behaviour of different types of sediment is critical if we are to design effective management plans. Cutting edge nuclear techniques have been developed to trace river silt back to source, in a jointly Coordinated Research Programme between the UN Food and Agriculture Organisation and the International Atomic Energy Agency. However, these new techniques have led to a step change in data complexity. While these datasets capture real world complexity in time and space, the conventional statistical approaches to quantify sediment provenance do not. This severely limits the power of the new tracing techniques. Advances in ecological source models based on Bayesian statistics, however, offer a solution. New models e.g. MixSIAR have been developed that can deal with the complexity in a quantitative way and, if tailored to specific river basin sediment data, will allow us to address the above challenge.

Research and Innovation Objectives:
1. To marry together the strengths of isotopic and biogeochemical sediment tracer technology in the EU with new developments in ecological source apportionment models led by US scientists to deliver a powerful sediment source apportionment toolkit to combat threats to global food and water security from soil erosion and siltation.
2. To demonstrate the capability of the integrated IMIXSED approach to deliver EU development policy goals in two demonstration catchments (Gilgel Gibe in Ethiopia and Manyara in Tanzania) where food, water and energy security is threatened by siltation of hydropower reservoirs.

The new IMIXSED approach to sediment tracing and apportionment of problem sediment in complex river basins offers better system representation by accounting for source contributions weighted by tributary. The step change proposed by IMIXSED in structuring data according to geographical catchment hierarchy leads to a more representative unmixing model structure. The approach offers a more accurate means of estimating the contribution of primary sediment sources linked to land use where more than one sub-catchment feeds a sediment problem downstream. The new tool for source apportionment in river basins offers wider application in systems affected by natural and human-induced change and the lessons learned are relevant to source apportionment applications across the world. The new data handling routines developed in IMIXSED support management of complex source apportionment challenges to help us tackle the global challenge of improved food, water and energy security through mitigation of soil erosion.
We demonstrated the capability of the integrated IMIXSED approach to deliver EU development policy goals in three demonstration catchments (Gilgel Gibe in Ethiopia, Manyara in Tanzania and Upper Chitlang in Nepal) where food, water and energy security is threatened by siltation of hydropower reservoirs. In the Ethiopian example, IMIXSED methodologies were applied in the catchment of the Gilgel Gibe hydroelectric dam, one of a series of development projects launched by the Federal Government of Ethiopia. The Gilgel Gibe reservoir is silting up and it has been proposed that landslides in the upper catchment may be an important sediment source. From a management perspective, the understanding of nature and relative importance of the principle sediment sources within a catchment is needed to support the design and implementation of sediment control strategies in catchments. The IMIXSED approach was applied in the Unta sub catchment using XRF geochemistry. Source discrimination indicated relative differences between the broad source groups but the main groupings separated surface and subsurface derived material. The mean contributions from individual source types (i.e. cultivated land, grass land, landslide, homestead and wood land) varied but landslides remained the dominant source to stream sediment because of their proximity to river (highly connected to river) and high abundance in the Unta catchment. To support land remediation and management change, further targeted research is now needed to develop knowledge of landslide processes and drivers in the landscape and the catchment-wide contribution of landslides to sediment delivery into the Gigel Gibe reservoir. In the Tanzanian example, work was based in the Lake Manyara basin in northern Tanzania, working in a subcatchment of the Makauni river catchment. Land use is mixed with widespread savannah grazing lands and agricultural land for crops. Sever gully erosion is observed in the grasslands. This part of the world is at risk of changing rainfall and runoff patterns with climate change and we hypothesised that land management factors will amplify the impacts of extreme rainfall events. Sedimentation threatens downstream ecosystem services in the Lake Manyara Man and biosphere UNESCO reserve. Research and stakeholder questions relate to erosion process and management solutions and we applied the IMIXSED approach to develop knowledge of sediment sources in different geomorphic zones of the study catchment. Sources were rationalised down to four groups: UPPER CATCHMENT forested, MIDDLE catchment surface and subsurface soil, LOWER catchment soil (undifferentiated between surface and subsurface since the soils here represent ‘recent’ reworked valley fill material). Results from the model indicated that while there is extensive visible gullying in mid catchment, surface erosion fed by sheet wash into existing gullies remains a key input. Fingerprinting reveals importance of sheet erosion on steeper land.
The IMIXSED work programme and network development has already supported successful applications for follow-on research funding e.g. a work programme funded by the UK Global Challenges Research Fund. A UK PhD studentship was also successfully awarded (based in UK partnered with Tanzanian and Belgian partners) to develop and upscale the Tanzanian research programme. The Chilean government has funded a Chilean PhD studentship to develop and test Deconvolutional MixSIAR in UK and Chilean settings. IMIXSED is contributing to a new Coordinate Research programme led by UN FAO and IAEA on “Nuclear Techniques for a Better Understanding of the Impact of Climate Change on Soil Erosion in Upland Agro-ecosystems (D1.50.17)” involving 12 participating countries including Morocco and Madagascar. The Tanzanian partners have also been invited to apply to join the new Regional African Technical Cooperation Project (2016-2020) on “Enhancing Regional Capacities for Assessing Soil Erosion and the Efficiency of Agricultural Soil Conservation Strategies through Fallout Radionuclides”. In terms of direct socio-economic impact, developing follow-on grants are laying the pathway for new evidence-based, bottom-up policy instruments in the Monduli catchment. IMIXSED data is a fundamental component of this process.
Gully erosion threatens agricultural land in Unta catchment, Ethiopia
IMIXSED team visit Tanzanian demonstration catchment
Laboratory capacity building
IMIXSED team visit Ethiopian demonstration catchment
IMIXSED team at NM-AIST
UK demonstration catchment training
Gully erosion in grasslands, Monduli catchment, Tanzania