Estimation of global river transport of sediments and associated particulate C, N, and P
This paper presents a multiple linear regression model developed for describing global river export of sediments (suspended solids, TSS) to coastal seas, and approaches for estimating organic carbon, nitrogen and phosphorous transported as particulate matter (POC, PN and PP) associated with sediments. The model, with river-basin spatial scale and a one year temporal scale, is based on 5 factors with a significant influence on TSS yields (the extent of marginal grassland and wetland rice, Fournier precipitation, Fournier slope and lithology), and accounts for sediment trapping in reservoirs.Earlier approaches in the literature did not incorporate anthropogenic factors as controls of TSS yields. We found a significant anthropogenic effect of cultivation (wetland rice extent, mainly occurring in Asia, and marginal grasslands which occur mainly in Asia and Africa). There is general agreement about the importance of climate, relief and lithology between all models. A further difference between our model and earlier ones is the inclusion of the effect of trapping of sediments in human-made reservoirs, which causes a global reduction of river loads of particulate nutrients of 13%.
Global river export of suspended solids and particulate carbon and nutrients
The global river export of TSS calculated thus is 19 Pg yr-1 with a 95% confidence interval of 11-27 Pg yr-1 when accounting for sediment trapping in regulated rivers. Associated POC, PN and PP export is 197 Tg yr-1 (as C), 30 Tg yr-1 (N) and 9 Tg yr-1 (P), respectively. The global sediment trapping included in these estimates is 13%. Most particulate nutrients are transported by rivers to the Pacific (~37% of global particulate nutrient export), Atlantic (28-29%) and Indian (~20%) Oceans, and the major source regions are Asia (~50% of global particulate nutrient export), South America (~20%) and Africa (12%).
We succeeded to develop a robust model for TSSY which was cross-validated by using training and validation data sets and validated against independent data. The model generates predictions within a factor of 4 for 80% of the 124 rivers in the data set. It is a robust model which was cross-validated by using training and validation sets of data, and validated against independent data. In addition, Monte Carlo simulations were used to deal with uncertainties in the model coefficients for the 5 model factors.
The model should not be used to predict the fluxes of sediments and particulate compounds for the mouth of individual rivers. They are more appropriate to estimate the regional, continental or global fluxes of sediment and particulate matter to the oceans, and changes therein as a result of climate and land-use change.
There is of necessity considerable uncertainty associated with the model’s predictions, and future efforts will no-doubt improve on our predictions significantly. Nonetheless, as the first attempt to develop a robust, internally consistent, and spatially-explicit global model of POC, PN, and PP river export, our model for particulate nutrients constitutes a significant advance in its own right.
|Author(s)||Beusen AHW ; Dekkers ALM ; Bouwman AF ; Ludwig W ; Harrison J|
|Publication||Global Biochemical Cycles 2005; 19:art.no. GB4S05|