A "thousand year event" means that --- according to the scientific data that we have accumulated over the past 200 years; plus a data base of all accounts of weather events that have been written down since writing was invented --- mathematicians have calculated a complex function that derives the probability that this or that weather event is likely to happen at this time and place. The result is a probability percentage.
Climate scientists use this formula to calculate the rarity (or not) of specific events by loading into this formula, weather data from the area in question over a specific period there ... and out pops the percentage chance of this sort of event happening there.
So in an event like that which happened in West Virginia June 23-24, 2016, it turned out to be likely to happen 0.001% (one, one thousandth) in any given year. So that adds up to the event probably happening once every thousand years. The percentage is predicated on the time spread of the data imputed (in the West Vierginia event described below, the meteorologist used a 30-year local data set spanning 1976 to 2006).
So ... as 1,000 year events happen more and more, climate scientists may be tempted to add data sets that include more recent data - and the algorithm will begin to pop out percentages that will reduce the improbability of these specific events happening - so this sort of events might slowly (based on what data set is used) become '500-year events'.
We have to be careful that we 'stay' the model at the point where this kind of weather started to become the new normal. And the new normal (for now) appears to be the kind of weather events that we're seeing this year - this year we have seen 3 of these 1000-year events in North America1,2,3 (so far).
I believe that Climate Scientists - starting now - should not feed current data points into the model to refine it's accuracy - keep using the pre-2016 data sets with regard to calculating the effects of global warming - because this is the Turning Point.
2016 is the turning point - the point at which the current record separates from the existing 'normal'.
This needs to be done so that over a person's life-time we will have a base-line from which to understand what is happening. So that in 20 years we will have within the popular zeitgeist, an understanding of the increase in 1,000-year events - and thus civilization as a whole - in the populist sense - will have the information to help them make informed choices (because we're obviously not making those kinds of choices now).
West Virginia 1000 year flood event - June 23-24, 2016
|Via Climate.gov 2016/07/08 West Virginia 1000 year flood June 23-24 - 24 hour accumulation map by Charleston, WV Forecast Office meteorologist, Nick Webb | https://www.climate.gov/sites/default/files/1000yr24hrRainfall_nwebb.jpg|
Ellicott City 1000 year flood event - July 30, 2016
|Baltimore Sun - 2016/07/31 | Ellicott City gets rainfall expected only once every millennium | http://www.baltimoresun.com/news/weather/weather-blog/bal-wx-ellicott-city-gets-rainfall-expected-only-once-every-few-hundred-years-20160731-story.html|
Louisiana 1000 year flood event - August 9-14, 2016
1 Climate.gov - 2016/07/08 | 'Thousand-year' downpour led to deadly West Virginia floods - by Tom DiLiberto | https://www.climate.gov/news-features/event-tracker/thousand-year-downpour-led-deadly-west-virginia-floods
2 Baltimore Sun - 2016/07/31 | Ellicott City gets rainfall expected only once every millennium | http://www.baltimoresun.com/news/weather/weather-blog/bal-wx-ellicott-city-gets-rainfall-expected-only-once-every-few-hundred-years-20160731-story.html
3 The Real News - 2016/08/19 | Devastating Floods in Louisiana Are the Sign of Things to Come | http://therealnews.com/t2/index.php?option=com_content&task=view&id=31&Itemid=74&jumival=17031