Back to Search
Start Over
An assessment of the accuracy of interpolated daily rainfall for New Zealand.
- Source :
- Journal of Hydrology (00221708); Jul2012, Vol. 51 Issue 1, p25-44, 20p, 2 Charts, 6 Graphs, 10 Maps
- Publication Year :
- 2012
-
Abstract
- Daily rainfall data from about 200 automatic weather stations are operationally interpolated onto a regular grid of points covering all of New Zealand at a resolution of 0.05 degrees lat/long (approximately 5 km). These data, as well as daily values for 10 other climate variables, are known as NIWA's Virtual Climate Station (VCS) data. There is a need to understand the accuracy of the daily rainfall estimates, particularly for hydrological assessments. The provision of historic daily rainfall data from more than 700 Regional Council locations (many of which are no longer operational) not used in the generation of the VCS estimates, has enabled a robust error assessment. Using a regression model approach (where the VCS rainfall is the independent variable and the Regional Council rainfall is the dependent variable), bias corrections were determined and applied to every VCS daily rainfall ≥ 1 mm. The bias-corrected VCS rainfall mean absolute error for locations below about 500 m elevation is approximately 2-4 mm (95% of the range) when rain days only (VCS rainfall ≥ 1 mm) are analysed. The error (and its spatial variation) is higher in areas above 500 m elevation, at approximately 5-15 mm. For daily VCS rainfalls > 40 mm (heavy rain days), the errors are approximately 8-12 mm (low elevation areas) and 10-40 mm (high elevation areas), and for monthly rainfall totals Journal of Hydrology (NZ) 51 (1): 25-44 2012 © New Zealand Hydrological Society (2012) the corresponding errors are approximately 10-15 mm and 10-120 mm. For some uses of daily rainfall data, the installation of good quality telemetered rain gauges is the only option. However, if gauges can't be installed or the data collection cannot be maintained long-term, then using the bias-corrected VCS rainfall data is a viable option that is often a significant improvement over using data from distant stations that may not be in the same catchment or even located in a similar climatic zone. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 00221708
- Volume :
- 51
- Issue :
- 1
- Database :
- Complementary Index
- Journal :
- Journal of Hydrology (00221708)
- Publication Type :
- Academic Journal
- Accession number :
- 78148471