Rain-shadow hydrology: Influences on river flows and flood magnitudes across the central massif divide of La Grande Terre Island, New Caledonia

Abstract

Within the tropical South Pacific islands, meaningful examination of how mountain barriers influence river flow characteristics has not previously been attempted, owing principally to the lack of reliable long-term data from an adequate number of basins that are geographically spread across opposing sides of a major topographic divide. The island of La Grande Terre in New Caledonia provides an exception, since robust flow records with effective durations between 15 and 52. years are available for 22 monitored rivers draining both sides of a massif dividing range. Cross-island differences in several fundamental hydrological parameters were statistically analysed, including runoff coefficients, median daily discharges and specific flood flows. Results identify that significant flow contrasts exist between windward versus leeward basins, which are not necessarily revealed by observing patterns of rainfall distribution over the island. Power relationships are derived using both empirical and theoretically-based models for maximum decadal (Q10y) and centennial (Q100y) flood size against catchment area, with basins grouped according to windward or leeward location on La Grande Terre. A key finding is that the rain-shadow influence (RI) is essentially constant (and therefore quantifiable) across the central massif divide. Flood magnitudes for both Q10y and Q100y events are between 35% and 80% greater for windward versus leeward basins of equivalent size, depending on whether empirical or accepted theoretical models are applied. The term 'rain-shadow hydrology' is proposed for the generalised concept of windward vs leeward divergence in hydrological patterns. In spite of being developed for the La Grande Terre situation, the theme has relevance for neighbouring high islands in the tropical South Pacific, and therefore contributes to understanding tropical island hydrological characteristics in the trade wind belt. However, the transferability of estimated ratios of the rain-shadow influence (RI ratios) requires validation with flow data from other islands, especially to establish windward vs leeward differences in flood magnitude-basin area relationships. © 2011 Elsevier B.V.