Estimating global river runoff from observed river flow data.

摘要

This thesis develops a new dataset of 981 observed river discharges to the ocean with a total discharge of 824,503 m3/s. The characteristics of this dataset are examined and a power law size distribution is developed for rivers between 250 m3/s and 20,000 m3/s. Using the observed data and the power law distribution to estimate missing data, a total global land surface runoff estimate is obtained that is in good agreement with previous estimates.; Past efforts and the methods used to estimate global land surface runoff are reviewed. The new global land surface runoff estimate of 1,256,640 m 3/s is within 2% of the mean value of 1,274,950 m3/s derived from 12 previous estimates. Similar estimates created for each continent differ from earlier continental estimates by up to 20%. The sum of the continental estimates is 1,124,940 m3/s, which is about 6% lower than the global estimate.