Quality control for daily observational rainfall series in the UK
Daily rainfall sequences in the UK need careful quality control before use. This issue is discussed using data from two rain gauge networks. In particular, changes in recording resolution or inconsistencies between different types of recording device may result in biases, for which allowance should...
Main Authors: | , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
BLACKWELL PUBLISHING
2006
|
Subjects: | |
Online Access: | http://discovery.ucl.ac.uk/86072/ |
Summary: | Daily rainfall sequences in the UK need careful quality control before use. This issue is discussed using data from two rain gauge networks. In particular, changes in recording resolution or inconsistencies between different types of recording device may result in biases, for which allowance should be made when these data are used for modelling. Model fitting is shown to be an effective way of investigating unexplained spatial or temporal inconsistencies. Generalized linear models are fitted to data from the Blackwater catchment to illustrate model-based data checking. Some apparently unexplained spatial inconsistencies are related to trace values in the data set. By applying a simple threshold to the data, the inconsistencies can be removed. |
---|