Regionalization of patterns of flow intermittence from gauging station records

摘要

Understanding large-scale patterns in flow intermittence is important foreffective river management. The duration and frequency of zero-flow periodsare associated with the ecological characteristics of rivers and haveimportant implications for water resources management. We used daily flowrecords from 628 gauging stations on rivers with minimally modified flowsdistributed throughout France to predict regional patterns of flowintermittence. For each station we calculated two annualtimes series describing flow intermittence; the frequency of zero-flowperiods (consecutive days of zero flow) in each year of record (FREQ; yr−1),and the total number of zero-flow days in each year of record (DUR; days).These time series were used to calculate two indices for each station, themean annual frequency of zero-flow periods (mFREQ; yr−1), and the meanduration of zero-flow periods (mDUR; days). Approximately 20% of stations hadrecorded at least one zero-flow period in their record. Dissimilaritiesbetween pairs of gauges calculated from the annual times series (FREQ and DUR) andgeographic distances were weakly correlated, indicating that there waslittle spatial synchronization of zero flow. A flow-regime classificationfor the gauging stations discriminated intermittent and perennial stations,and an intermittence classification grouped intermittent stations into threeclasses based on the values of mFREQ and mDUR. We used random forest (RF) models torelate the flow-regime and intermittence classifications to severalenvironmental characteristics of the gauging station catchments. The RFmodel of the flow-regime classification had a cross-validated Cohen's kappaof 0.47, indicating fair performance and the intermittence classificationhad poor performance (cross-validated Cohen's kappa of 0.35). Bothclassification models identified significant environment-intermittenceassociations, in particular with regional-scale climate patterns and alsocatchment area, shape and slope. However, we suggest that the fair-to-poorperformance of the classification models is because intermittence is alsocontrolled by processes operating at scales smaller than catchments,such as groundwater-table fluctuations and seepage through permeablechannels. We suggest that high spatial heterogeneity in these small-scaleprocesses partly explains the low spatial synchronization of zero flows.While 20% of gauges were classified as intermittent, the flow-regimemodel predicted 39% of all river segments to be intermittent, indicatingthat the gauging station network under-represents intermittent riversegments in France. Predictions of regional patterns in flow intermittenceprovide useful information for applications including environmentalflow setting, estimating assimilative capacity for contaminants, designingbio-monitoring programs and making preliminary predictions of the effects ofclimate change on flow intermittence.