Alain Poirel

pH modelling by neural networks. Application of control and validation data series in the Middle Loire river

Artificial neural networks (ANNs) are applied as a new type of model to estimate the daily pH of the Middle Loire river. The model is used for pH measurement screening, error detection (abnormal values, discontinuities and recording drifts) and validating the collected data. The measured values of pH are compared with the values estimated by the ANN model using statistical tests to verify homogeneity and stationarity. River water pH is affected by numerous processes: biological, physical and geochemical. Examples are: CO2 pressure equilibrium with the atmosphere, photosynthesis, respiration of plants, organic matter degradation, geological and mineral background, pollution etc. Inter-relationships between these processes and pH values are complex, non-linear and not well understood. As a neural network provides a non-linear function mapping of a set of input variables into the corresponding network output, without the requirement of having to specify the actual mathematical form of the relation between the input and output variables, it has the versatility for modelling a wide range of complex non-linear phenomena. For this reason the neural network approach has been selected and tested for pH modelling. We used the classical multilayer perceptron model (MLP). River discharge and solar radiation variables are used as inputs to the MLP model. The choice of these variables is dictated by what are perceived to be the predominant processes that control pH in the Middle Loire river, which is typically eutrophic during the low-flow summer period. The influence of the previous day’s flows and radiation has been evaluated in the calibration and verification test. The best network found to simulate pH was one with two input nodes and three hidden nodes. The inputs are: daily discharge and a variable called ‘Index of anterior radiation’, i.e. calculated as an exponential smoothing of the daily radiation variable. When calibrated over 4 years of data and tested (i.e. verified) for a one-year independent set of data, the model proved satisfactory on pH simulations, with accuracies in the order of 86%. After elaborating the pH model, the Student test and the cumulative Page‐Hinkley test were applied for automatic detection of changes in the mean of the residuals from the ANN pH model. This analysis has shown that such tests are capable of detecting a measurement error occurring over a short period of time (1‐4 days).

A quality-control method for physical and chemical monitoring data. Application to dissolved oxygen levels in the river Loire (France)

Abstract A quality-control method is proposed for examining continuous physical and chemical measurements, including temperature, dissolved oxygen, pH and electrical conductivity. Firstly, measurement consistency is evaluated by various modelling approaches: internal series structure, inter-variable relations or relations with external variables, spatial coherence and deterministic models. Secondly, outliers or systematic errors are detected using classical statistical tests. The method was evaluated for dissolved oxygen concentrations (DO) in the river Loire at Dampierre over a 5-year period (1990–1994), using data records containing fictitious errors, and raw data for the year 1995. The results demonstrate the effectiveness and advantages of a multi-model approach. In the case of dissolved oxygen for example, slow continuous drifts are always detected in under 4 days.

Badlands as a hot spot of petrogenic contribution to riverine particulate organic carbon to the Gulf of Lion (NW Mediterranean Sea): POC flux from badlands to the NW Mediterranean sea

Determining the riverine carbon fluxes to oceans is critical for an improved understanding of C budgets and biogeochemical cycles (C, O) over a broad range of spatial and time scales. Among the particulate organic carbon (POC) involved in these fluxes, those yielded by sedimentary rocks (petrogenic POC: pPOC) remain somewhat uncertain as to their source on continental surfaces. Based on time series from long‐term observatories, we refine the POC and sediments flux of the Rhone River, one of the major tributaries to the Mediterranean Sea. Radiocarbon measurements on a set of riverine samples and forward modelling were used to (i) determine a modelled pPOC content and pPOC/POC ratio for each sample set, (ii) assess pPOC flux delivered to the NW Mediterranean Sea, and (iii) estimate the badlands contribution from the Durance catchment to both the pPOC and to sediment discharges. The weighted pPOC flux contributes up to 26% of the POC flux (145 Gg yr‐1) discharged into the Mediterranean Sea, whereas the weighted pPOC content reaches 0.31 wt%. Despite their low contributive surface area (0.2%), badlands provide, respectively, 12, 3.5 and 14% of the pPOC, POC and sediment fluxes to the Rhone River. Consequently, such rocks can be considered as a major source of pPOC and sediments for the NW Mediterranean Sea and potentially for oceans. We suggest that river‐dominated ocean margins, such as the Rhone River, with badlands in their catchment could export a significant amount of pPOC to the oceans.