San Francisco just had its second-rainiest day since 1849

Is there data for cyclonic rainfall that might be modeled for similar to how this article titled “How can we call something a thousand-year storm if we don’t have a thousand years of climate observations” is?

http://www.climate.gov/news-features/event-tracker/how-can-w...

None of that detracts from this of course.

This is one thing that surprised me when moving from NY (lots of snow) to SF.

Rain in SF seems to trigger traffic conditions similar to light snow or ice in the northeast.

I’ve always been curious what causes this. (My guess is people are slower to replace balding tires when there isn’t frequent rain or snow)

Plotting winter rainfall does indeed show a small trend towards wetter winters (r² = ~0.1). But the plot looks odd. Virtually all the trend comes from a step change starting around 1980. Since then the trend has been tiny (r² = 0.02). The trend from 1860 to 1980 is flat (r² = 0.0018). This doesn't look much like what you'd expect from a continuous process, which CO2 emission has been:

https://gml.noaa.gov/ccgg/trends/

This sort of sudden discontinuity raises the question of whether there was a methodology change at some point. A quick check shows that there was:

https://www.metoffice.gov.uk/weather/guides/observations/how...

"As automated instruments were introduced across the synoptic network in the 1980s and 1990s the 5 inch gauge was still deployed alongside the tipping bucket gauge to continue a long consistent record of measurements for climate purposes. In recent years this practice has proved impractical and many automatic sites now only report rainfall amount from a tipping bucket gauge."

So around the time they start introducing automation to the network and switch to tipping bucket gauges, winter rainfall measurements go up a bit and then re-stabilize. Hypothesis: the old network that relied on manual measurements by volunteers was prone to under-measure in winter when people don't want to go outside.

What about summer rainfall? The cited atmospheric effect would presumably hold regardless of season as it's just about warmer air = more moisture in it. But here we see there's actually a slight downward trend, although yet again it's so trivial as to be irrelevant (r² = 0.0036). There is no change starting in the 1980s either, which would weakly support the idea of a problem with prior winter under-collection.

So it looks like not much is happening in this data. To sanity check that I grabbed a Met Office report on rainfall trends. It says:

"An analysis of annual, summer and winter rainfall observations by Jenkins et al. (2008) shows that annual mean precipitation over England and Wales has not changed significantly since 1766 when the records began"

https://www.ofwat.gov.uk/wp-content/uploads/2015/11/rpt_com_...

If global warming is changing rainfall patterns it isn't having much effect, and to the extent small trends in the winter data do exist it's possible that the data set has been affected by methodology changes. Note that if such a problem does exist the common practice of only analyzing data since 1960 would yield a maximally biased trend, as the long period of absolute stability up until 1980 would largely be erased.