Are transit times useful process-based tools for flow prediction and classification in ungauged basins in montane regions?

Transit times are being increasingly explored as process-based tools for conceptualizing hydrological function at a range of scales. Despite this, little effort has been made to relate transit times to conventional hydrometric flow statistics. Rather, the identification of the appropriate transit time distribution (TTD) for a hydrological system and the derivation of metrics such as the mean transit time (MTT) have required quantitative assessment of input-output relationships for conservative tracers using lumped parameter models. This has allowed the main landscape controls to be identified and has facilitated the prediction of MTTs in ungauged basins in particular geomorphic provinces, though relationships with streamflow measures have been unexamined. We used estimated MTTs (with uncertainty) for 16 experimental catchments (0.5-690 km(2) in area) with contrasting geologic, topographic, pedologic and climatic characteristics in Scotland. The MTT was highly variable ranging from 60 days to ca > 1500 days, reflecting differences in catchment soil cover, geomorphic properties and precipitation. The MTT was closely correlated with key hydrometric design statistics such as the mean annual flood (MAF) and percentiles of high (Q(5)) and low (Q(95)) flows. Analysis of MTT estimates, in conjunction with geographic information system (GIS)-based assessment of landscape controls, showed that MTT could be predicted to within 30% for ungauged basins from catchment soil cover and drainage density. Furthermore, hydrometric design statistics for ungauged basins could also be forecast from MTT predictions with median relative errors of 14, 11 and 28% for the MAF, Q(5) and Q(95) respectively. We suggest that MTTs-predicted from mapped landscape characteristics-can be useful diagnostic metrics for ungauged montane basins, as well as a useful similarity index for process-based catchment classification. This is important as montane headwaters are often critical, but data-poor environments influencing the quantity, quality and ecology of downstream flows.