Elsevier

CATENA

Volume 38, Issue 3, January 2000, Pages 191-209
CATENA

The impact of rainstorms on floods in ephemeral channels in southeast Spain

https://doi.org/10.1016/S0341-8162(99)00071-5Get rights and content

Abstract

This paper presents and discusses data on rainfall, stage and estimated discharge for a large flood occurring in two catchments in southeast Spain in September 1997. Rainfall and stage were recorded using automatic logging equipment and discharge was estimated using measurements of channel cross-sections and water depth estimated from trash lines. Total precipitation in the Rambla de Torrealvilla was 50 mm in 2 days with maximum rainfall intensities of 80 mm h−1. Total rainfall in the Rambla de Nogalte was 195 mm in 3 days, with maximum intensities of 200 mm h−1. In the Torrealvilla, this rainfall produced three flood peaks with maximum stage approaching 2.5 m. In the Nogalte, there was only one flood peak, which was 0.5 m deep. Estimated discharge varied widely throughout both catchments with maxima of 120 m3 s−1 in the Torrealvilla and 60 m3 s−1 in the Nogalte. Maximum discharges occurred at times of high rainfall intensity, but intensity alone did not explain why some tributaries had very small discharges. Variations in discharge in the ephemeral channels were due to combinations of lithology, morphology and land use. The predominantly marl catchment of the Torrealvilla had a lower threshold rainfall intensity than the schists of the Nogalte. Within each catchment sub-basins characterised by steep, gorge like terrain and sub-basins where agriculture had been abandoned both resulted in higher flood discharge. The contributing areas for the September storms were up to two thirds of tributary catchment areas. Comparison of rainfall data records shows that the September flood was the fifth largest on record and had a recurrence interval of 7 years. The largest (1973) flood, which is known to have caused substantial damage and a number of deaths, was only a 30-year event. The floods on the Torrealvilla destroyed at least two check dams and evidence suggests that these had little effect on reducing the impact of the floods.

Introduction

The conditions, position and connectivity of runoff and sediment source areas in semi-arid catchments have a profound effect on flood characteristics of ephemeral streams. The relationship between rainfall, the source areas of runoff and the flood hydrograph is complex and, because of the discrete nature of each convective cell, an individual storm is unlikely to affect the entire drainage basin and successive storms will affect different parts of a catchment (Schick and Lekack, 1987). Total annual precipitation for the region is less then 500 mm, and this is concentrated in a few winter months. Rainfall in the Mediterranean is usually very intense (Obled and Tourasse, 1994; IGN, 1995), exhibits an erratic temporal and spatial distribution, and can be very localized (Thornes, 1976; Alonso-Sarria and López-Bermúdez, 1994). The likelihood of very large floods is therefore small because the precise conditions needed to produce outflow from sub basins across the whole of a catchment rarely occur.

This paper presents partial records of rainfall and runoff for floods in two catchments in southeast Spain in September 1997. The recorded rainfall data are linked to estimates of peak flood discharge that were made throughout the catchments following the storms. These data are used to investigate variations in discharge throughout the ephemeral channel systems and to estimate rainfall intensities and return periods for previous large floods. These estimates of changing flood discharge are some of the first of their kind and allow us insight into conditions promoting floods.

Section snippets

Study area

The work was carried out in two catchments; the Rambla de Nogalte located on the border of the Provinces of Murcia and Almerı́a, and the Rambla de Torrealvilla located in the Province of Murcia, southeast Spain (Fig. 1). Characteristics of the two study basins are summarised in Table 1. The Rambla de Nogalte is the larger basin and drains mainly metamorphic rocks and conglomerates, dominated by red mica schist, but with localised outcrops of blue mica schist. The soils derived from the

Nogalte

Fig. 2 shows rainfall intensities from 11 March 1997 to 5 January 1998. The floods in the Nogalte were caused by the storms on the 29 September 1997. For this day, all the rain gauges registered three intense rain events. At most rain gauges, the first two storms had similar maximum intensities, approaching 200 mm h−1 and the third storm was much less intense at approximately 50 mm h−1. The maximum rainfall intensities were much higher than expected.

These data can be plotted at a much greater

Variations in discharge due to lithology

The argument that flood discharge is influenced by lithology is supported by the difference in lithology and discharges between the two study catchments, but also by variations locally within the basins. Differences in lithology between the two catchments results in a lower threshold rainfall for the Torrealvilla than that for the Nogalte. Initial estimates of the thresholds (as calculated above) are 50 and 150 mm h−1, respectively. Flood discharge is also influenced by transmission losses,

Conclusions

The data record from the rain gauges in the Rambla de Nogalte catchment showed that the network is too sparse to accurately relate rainfall to discharge. The difference in timing between flow and rainfall at the stage gauging site and rain gauge S1 showed how rainfall varies over distances as small as 2.5 km. Thus, the intense storm cells that result in flows in semi-arid channels may be much smaller than estimated by Renard and Keppel (1966)and Diskin and Lane (1972). If rain gauges were going

Acknowledgements

This work has been carried out as part of the MEDALUS III project. We would also like to thank Mark Newcombe in the Graphics Unit in the School of Geography, University of Leeds for Fig. 1, Fig. 3, Fig. 6, Dr. Roy Richardson for helping install the rain gauges and Fernando Garcia Robredo for the 30-year data record. Automatic monitoring equipment was funded by King's College London in the Nogalte catchment and by Leeds, Portsmouth and Leuven in the Torrealvilla.

References (17)

  • Ackers, P., White, W.R., Perkins, J.A., Harrison, A.J.M., 1978. Weirs and Flumes for Flow Measurement. Wiley,...
  • Acrement, G.J., Schneider, V.R., 1990. Guide for selecting Manning's roughness coefficient for natural channels and...
  • F. Alonso-Sarria et al.

    Rainfall time and space variability during short storms in southeast Spain

    Geökodynamik Band

    (1994)
  • Barnes, H.H., 1967. Roughness characteristics of natural channels. US Geol. Survey Water Supply Paper...
  • Conesa-Garcia, C., 1995. Torrential flow frequency and morphological adjustments of ephemeral channels in South East...
  • M.H. Diskin et al.

    A basinwide stochastic model of ephemeral stream runoff in south-eastern Arizona

    Int. Assoc. Sci. Hydrol. Bull.

    (1972)
  • Frostick, L.E., Reid, I., 1979. Drainage-net control of sedimentary parameters in sand bed ephemeral streams. In:...
  • W.H. Green et al.

    Studies in soil physics: 1. The flow of air and water through soils

    J. Agric. Soils

    (1911)
There are more references available in the full text version of this article.

Cited by (101)

  • Flood Generation and Flood Waves

    2022, Treatise on Geomorphology
View all citing articles on Scopus
1

Tel.: +44-171-873-2738.

2

Tel.: +44-1705-842-589.

View full text