ISSN:
1475-2743
Source:
Blackwell Publishing Journal Backfiles 1879-2005
Topics:
Geosciences
,
Agriculture, Forestry, Horticulture, Fishery, Domestic Science, Nutrition
Notes:
Abstract. Estimates of winter drainage, D, are available for 82 locations in England and Wales for each winter during the 35-year period, 1940/1 to 1974/5. At each location, D is approximately normally distributed. The variability, expressed as a coefficient of variation, is approximately inversely proportional to the average drainage; it is large, ranging from 20–30% in Wales and the uplands of northern and western England to 60–70% in eastern England. Assuming a normal distribution, the probability of exceeding a given amount of drainage is easily found.The proportion, f of solute initially in the soil that is leached each winter can be calculated using these estimates of D in a simple layer model. However, f is not normally distributed, but can be fitted by the Type-1 extreme-value distribution.If there are n years of data, and m is the rank of a value in order of decreasing size, the estimation of the probability, P, of exceeding the mth largest value, fm, is P(ffm) =m/(n+1), and the return period, t, is 1/P years. Using the transform, W= In[—In P], graphs of/against W yield families of lines that, for a given location and depth in the soil, depend only on soil type, and from which extreme values likely to be encountered in a given return period can be forecast. Results are discussed for Mildenhall, Suffolk (= 99 ± 68 mm), Cockle Park, Northumberland (±= 275 ± 125 mm), Welshpool, Powis (±= 373 ± 121 mm), Carmarthen, Dyfed (±= 764 ± 196 mm) and Ambleside, Cumbria (±= 1280 ± 276 mm).
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1111/j.1475-2743.1991.tb00859.x
Permalink
