THEME 3. LOSS OF WATERSHED FUNCTIONS

Posters on Watershed Functions

Erosion in Bench Terraced Upland Volcanic Terrain, West Java, Indonesia

by Albert I.J.M. van Dijk and LA. (Sampurno) Bruijnzeel¹, and Edi Purwanto²

II. Process modelling across a range of scales.

Results obtained during the first phase of the Cikumutuk Hydrology and Erosion Research Project (CHERP) suggest that the bulk of the sediment produced in this volcanic upland agricultural catchment originates from (the bare risers of) rainfed bench terraces, particularly on steep slopes. Also, a surprisingly low sediment delivery ratio was obtained for a sub-catchment draining c.180 terraces (see Part I). To clarify the underlying processes, the second phase of the project (1998-2002) includes an effort to model these processes in a physically meaningful way across a continuum of scales, i.e. from individual terrace risers and beds, via the toe drains, to drainage ways along field bound-aries into gullies, and finally into the stream draining the entire catchment. Three physically-based models are used: (i) VAMPS (Schellekens, 1996) for modelling the one-dimensional water balance of a vegetated surface; (ii) GUEST (Rose, 1993) to model on-site runoff rates, erosion and sediment transport; and (iii) TOPOG (Vertessy et al., 1990) to model (sub-)catchment hydrology, erosion and sediment transport in a spatially distributed manner. The three models will be integrated and modified where necessary to be applicable in the given setting (backsloping bench terraces). Examples of modifications include:

• incorporate direct sediment transport by splash in GUEST and allowing for lateral contributions from both bed and riser sections to the central toe drain;

• develop a method to translate DEM/TOPOG-based slope data into terrace geometry data (size and slopes of riser, bed and toe drain) for use in GUEST, e.g. through empirical relationships between slope steepness and terrace geometry based on field surveys;

• develop and incorporate a procedure to model runoff and sediment dynamics in drainage ways.

To validate the models and their modifications, field data are being collected in addition to the baseline monitoring programme (cf. Part I), including:

• continuous measurements of water and sediment outputs from terrace riser and bed sections as well as from complete terrace units of variable geometry;

• event-based separated measurements of splash and wash erosion;

• monitoring changes in sediment storage at the respective levels of scale.

The final result (expected around 2002), will be an integrated and validated combination of the above-mentioned models and their modifications which will enable the prediction of runoff, erosion and sediment transport in tropical terraced terrain for soil conservation and watershed management purposes, in a user-friendly but physically meaningful way.

Footnote_______________

¹ Free University, Faculty of Earth Sciences, Department of Environmental Hydrology, De Boelelaan 1085, 1081 HV Amsterdam. Telp: 31 20 4447294; Fax: 31 20 6462457; Email: brul@zorba.geo.vu.nl

² Centre for Education and Training of Forestry and Plantation Officials, Bogor, Indonesia