Report of Working Group I:

Leicester meeting 18-21 April 1999

Workgroup 1 - Linking processes across temporal and spatial scales

The first meeting of WG1 was held at Leicester as part of the COST/BGRG meeting 18-21 April 1999. The primary objective of this meeting was to set the programme of meetings for WG1 for the next 2-3 years of the COST action, in the light of discussed and agreed scientific priorities. The discussion was generated from a number of broad-ranging presentations, designed to spread across the full range of the working group's interests. In addition one particular issue was explored in greater depth, in association with the BGRG (British Geomorphological Research Group) working group on Rainfall Simulation. The following were formally invited on behalf of WG1:

14 presentations were made at the meeting for WG1:

Session 1: Rainfall Simulation (18-19 April)

John Quinton (Silsoe) and Dave Favis-Mortlock:
Rainfall simulators and erosion modelling
Dwayne Edwards (Kentucky): Invited Speaker.
Construction and performance of a plot-scale, fixed nozzle rainfall simulator.
Dave Favis-Mortlock (Oxford) and Tony Guerra :
Spatial variability in a rainfall simulation experiment: effects of different rainfall intensities

Chair : Adolfo Calvo (U. Valencia)

Session 2: Scale Issues: (19-20 April)

John Thornes (KC London):
Aggregation and Dis-aggregation for data and models
F. Darboux (1, 2), C. Gascuel-Odoux (1), P. Davy (2) & A. Crave (2); (1) Science du sol - INRA - Rennes; (2) Geosciences - UPR A4661 CNRS - Rennes:
Roughness and characteristic length of overland flow
Veronique Auzet (Strasbourg):
Implications of spatial heterogeneity for erosion modelling.
Mike Kirkby (Leeds):
Reconciling models across time and space scales:
Aaron Yair (Hebrew Univ, Jerusalem): Invited Speaker.
Spatial variability in the frequency and magnitude of runoff in a small arid watershed: Implications for water redistribution under changing climatic conditions
Anton Imeson (Amsterdam):
Nesting of Models
Keith Beven (Lancaster):
Mapping landscape spaces into model space
Mark Macklin (Aberystwyth):
Sediment linkages in flood plain systems
Adam Kertesz (Budapest):
Scaling up from plot to catchment in Hungarain catchments.
John Wainwright, Nick Drake and Q. Zhang (KC London):
Global scale modelling for soil erosion.

Chair Mike Kirkby (Leeds)

During parallel meetings of all three WGs, the following attended the WG1 session:

Summary of scientific discussions:

In environmental sciences, numerous processes act and feedback loops make interactions complex. Moreover, processes responsible for soil erosion have different spatial and temporal scales. For practical and economical limitations, results of a study carried out at one scale have to be transferred to others. Such extrapolation is not straightforward because system behaviour is non-linear and variable across scales (as also discussed in WG2).

Basically, a process is dominant over a range of spatial and temporal scales. Origin and prediction of emergent processes were questioned. Downscaling and upscaling can be performed thanks to recognition of similar pattern or with the use of general laws. Nevertheless, no general method for scale changes exists. To transfer data/knowledge across a range of scales always need hypothesis. For upscaling purpose, a simplification of representations have to be performed. In some cases, formal links could be drawn, based on fundamental physical concepts. The recognition of similar object at different absolute scale make easier extrapolation both for up- and downscaling.

In order to understand effects of current modifications and to draw causality links, extrapolation from past-climate changes can be performed. Nevertheless, such approach is limited by the conservation of natural archives and non-accessible parameters.

Illustrated using rainfall simulations, scale change appears feasible. Nevertheless, extension of relationships out of a defined context is uncertain. Spatial structure could be a leading factor in order to explain discrepancies from site to site. Both current environment variability and climate change heterogeneity have to be considered for soil erosion predictive studies. Hence, a pattern classification could potentially improve and fasten predictions and scale transfers.

Generally speaking, to validate scale changes remains a difficult task (see WG3).

Proposed schedule of future WG1 meetings

Topic Status Other WG's Location Organiser Date
Soil Erosion, SOM & Soil Biology (1GCTE Session together with COST sessions on The role of biological factors for water and wind erosion) C 2 Reading Kirkby Sep-99
Calibration, validation & Uncertainty in models at different scales P 2, MC Leuven Beven Apr-00
Linkage of hillslope erosion to sediment transport and storage in river and flood-plain systems P   Spain or UK ?? Oct-00
1 Processes of runoff Generation P   Israel Yair Apr-01
1 Surface Characteristics P 2 France Valentin Sep-01
1 Processes of sediment production and transport M   Spain or UK ?? Apr-02
1 Spatial and temporal structures and their evolution M   Germany Helming May-03

C=Confirmed, P=Provisional, M=Still at early planning stage
1=Four meetings under title of Runoff variability and its spatial structure. Each workshop contains all four themes, but with a different Lead Theme

These meetings, together with those for other working groups, were presented to participants in the closing plenary session of the Leicester Meeting, and approved in principle by the Management Committee. Fuller details will be provided as the programme of each meeting develops.

The meeting in Leicester was a successful initial meeting for COST623. Clearly there are lessons to be learned about the most effective way to administer future meetings through COST, and new proposals have been made through the Management Committee to achieve wider participation from all member countries.

Véronique Auzet
Mike Kirkby
1 May 1999
Last updated 20/11/1999 17:10:39