Cumulative Rainfall Review

Background

Daily rainfall totals can be compared to long term averages. Daily totals will vary depending on prevailing weather conditions and will show wetter or dryer periods than average.

This routine sets out to present trends around the long term average over a Jul-Jun 12 month period (southern hemisphere Winter-Spring-Summer-Autumn-Winter period). This is useful in assessing likelihood of droughty conditions if prevailing rainfall is trending low from the beginning of a period.

Description

Two scripts are provided here. The first, DailyRainfall_CreateReferenceData_10yrs.R, creates reference data for a user-specified time period (10 years is used in the script). This reference data is stored locally on a file share and then re-used in the second script, DailyRainfall_Accumulation_Outputs.R for comparison against daily rainfalls and the output of charts for each rainfall station.

DailyRainfall_CreateReferenceData_10yrs.R's function is to:

1. Discover those sites with the requisite measurements contained within the referenced file based on the collection name
2. Determine which of those sites has requisite period of record
3. Retrieve daily rainfall records for that period
4. Rearrange the data from a Jan-Dec to a Jul-Jun year
5. Calculate cummulative rainfall totals
6. Calculate cummulative rainfall percentiles
7. Save as reference data

DailyRainfall_Accumulation_Outputs.R function is to:

1. Load the reference data
2. Retrieve daily rainfall for current and previous year
3. Munge data for plotting
4. Plot

• Cummulative rainfall totals
• Compare cummulative rainfall to median and previous year
• Monthly rainfall totals

Note

1. Scripts are written for data in a Hilltop Server environment. For other time series servers, you will need to ammend

• the arguments used in the URL's to query the data
• the XPATH used to extract data from the XML responses

2. You will need to modify most of variables specified at the top of the scripts to match your local enviroments.

Acknowledgements

Thanks to the Catchment Data team of Horizons Regional Council for review and helpful comments in the development of this script. Many thanks also to the creators and maintainers of the XML, reshape2, zoo and hydroTSM library's.