Toepassing van de kansberekening op den regenval met het oog op polderbemaling

Toepassing van de kansberekening op den regenval met het oog op polderbemaling

Visser, B.

Tienstra, J.M. (promotor)
Bottema, O. (promotor)

Civil Engineering and Geosciences

1947-04-23

This study aims at aiming an insight into the rainfall in the Netherlands with regard to the determination of the desirable capacity of draining-mills. With this end in view a statistical investigation has been made of the data of the daily quantities of rain during the years 1890 up to 1939 inclusive, at Boskoop, Leeghwater and Hoorn. First of all we have made a division into days with a rainfall up to 5 mm. and days with a rainfall exceeding 5 mm. In various ways we have tried to find out whether the probability of the rainfall on a particular day being at least 5.1 mm. is independent of the rainfall on the other days. The precipitations (classified: up to 5 mm. and exceeding 5 mm.) turn out not to be independent. Next each period of one or more successive days with a rainfall exceeding 5 mm., preceded and followed by a day with a precipitation up to 5 mm., has been considered as a unity, the remaining single days with a rainfall up to 5 mm. being also considered as unities. The aggregate of these two kinds of unities, e. g. for 50 months of January has been considered as a population. The observations during the months of January 1890, January 1891 etc. up to 1939 inclusive have been considered as 50 samples. The hypothesis that these two kinds of unities are independent has been tested and proves to be in accordance with the samples. Finally the distribution of the rainfall over the various days of the periods of days with a precipitation exceeding 5 mm., e. g. for 50 months of January, has been examined; the amounts are classified according to the division given in the introduction. The daily amounts of precipitation during these periods; are independent. Consequently (and this is very important) the mean daily precipitations for periods (of days with a rainfall exceeding 5 mm.) of various duration are equal for a series of months of the same name. The author has shown how, with the aid of the investigations mentioned, probabilities of future periods can be calculated. From the observations we can infer a frequency distribution, the irregularities Of which are ascribable to chance. It is necessary to smooth this irregular distribution. Pearson and Goodrich give suitable methods, which have been compared. Goodrich's method proves to lead to the end in view quicker than Pearson 's; for the rest the two methods are equally serviceable. Finally some practical problems arising in polder-draining have been formulated and solved. Attention has been specially drawn to the mean daily precipitation during periods of days with a rainfall exceeding 5 mm.

rainfall
drainage
statistics
polder

http://resolver.tudelft.nl/uuid:74dafe7c-eb8a-4cf5-9199-253a85075996

IJdo Leiden

Institutional Repository

doctoral thesis

(c) 1947 Visser, B.