Key controls on water transit times along a tropical precipitation gradient

Abstract

The tropics are one of the most diverse and dynamic regions on Earth. Despite their importance, our understanding of tropical hydrological processes remains a significant challenge, mostly due to limited monitoring. Here, we used high-resolution daily input–output isotope data sets from seven tropical catchments ranging in size from 1.6 to 990 km2, situated within a precipitation gradient ranging from on average 958 to 5117 mm year−1, in Australia, Costa Rica, and Ecuador. For these catchments, we estimated and compared streamflow mean transit times (MTT) with potential explanatory hydro-geomorphological and climate variables. Using a simple lumped convolution integral model with a Gamma distribution as transfer function (best-fit Kling Gupta efficiencies up to 0.92), our tropical dataset resulted in short MTTs from 49 to 497 days (0.13 – 1.4 years), together with considerable presence of new water in the stream and low damping ratios in most of the catchments. The gamma distribution alpha parameter was below the previously identified quasi-global pattern of around 0.5 in 5 out of 7 catchments. Principal Component Analysis (PCA) and a relative importance test identified rock type (%), days with zero precipitation and the water storage capacity as the most important controls on TT. In addition, the TT distribution as indicated by the Gamma distribution alpha parameter was best explained by annual evapotranspiration, altitude, and Andosol soil cover (%). Our findings identified the key controls on TT and TT distribution in fast-responding tropical catchments compared to other geomorphic and climate zones, emphasizing the value of TT as a coherent descriptor that can be used for catchment characterization.