XRain extreme precipitation data is based on satellite precipitation measurements (more info), available almost anywhere across the globe. However satellite precipitation data is by nature less accurate than data derived from rain gauges, and therefore we recommend that you calibrate XRain statistics against local data where possible.

Scenario 1: You have some information at your location of interest

Let’s say you have 24 hour depths for your site that were derived from a rain gauge record, but you need to know sub-daily durations too.

In this scenario, we recommend that XRain data is scaled to match the available local data. Calculate the ratio between local data depths and XRain depths, and apply that ratio to XRain depths at other durations for the same ARI.


Daily totals have been recorded at a particular rain gauge for some time, and from this it is calculated that the 24 hour 20 year ARI is 101 mm. At the same location XRain estimates the following depths for the 20 year ARI:

1 hr2 hrs6 hrs12 hrs24 hrs
37.6 mm60.7 mm95.3 mm106 mm116 mm
XRain estimates

Of note, the XRain 24 hour 20 year ARI depth is 116 mm. The ratio between local data and XRain data at this ARI is therefore 101/116 = 0.87. We can therefore calibrate XRain depths in the above table by multiplying each value by 0.87, giving the following:

1 hr2 hrs6 hrs12 hrs24 hrs
32.7 mm52.9 mm83.0 mm92.3 mm101.0 mm
Calibrated XRain estimates
Scenario 2: The nearest local information is some distance away from your location of interest

The “Validation” section of the XRain report contains a comparison between XRain and NOAA Atlas 14, a USA service based on rain gauge data. This comparison (see Figure 6) shows that the percentage differences are reasonably consistent spatially; i.e. they don’t vary very quickly. As a result, it makes sense to scale XRain data based on the differences seen at nearby sites.

Therefore, if extreme precipitation data is not available at the location of interest but is available at a neighboring site, we recommend the following:

  1. Obtain XRain estimates at the neighbouring site for the same durations and AEP or ARI values. Calculate the ratio between supplied estimates and XRain estimates for each duration and AEP or ARI.
  2. Obtain XRain estimates at the location of interest and multiply these by the ratios calculated at the neighbouring site.

This strategy is based on the assumption that percentage differences are reasonably consistent spatially. Of course, the closer the calibration site is, the more confident we can be.

In both scenarios it may be advisable to add a factor of safety to estimated values.

Please note: due to known issues with GPROF microwave estimates used in the underlying dataset, we do not recommend comparing data over water bodies to data over land.