Long Range Weather and Climate Forecasting? Building for distant future
March 23, 2021 7:48 AM   Subscribe

Is there a source that makes localized forecasts of weather and climate in the far (years to decades) future and presents them in a unified, data driven presentation? There are lots of broad models of global and regional changes presented in narrative form, but is there something that could, say, take an address and present a forecast or prediction of future climate?

This is obviously fraught with uncertainty. Even a range might be helpful. Surely there are well-funded civil and infrastructure projects that do this work at least on an ad hoc basis. How does that work? Hire an climate scientist to write a report? If one wants to build a project that will last more than a generation or two (or more) how do we evaluate long rang climate effects? What's the market and technology for a public-oriented or publicly-accessible approach?
posted by GPF to Science & Nature (9 answers total) 3 users marked this as a favorite
I don't think this should be that difficult. Look up recent data, find the outliers, and make an assumption they are going to become more common for whatever criteria you are looking for. In terms of constructing a building, that's what you should do.
posted by The_Vegetables at 7:54 AM on March 23

Not by address, but the University of Maryland Center for Environmental Science has done this for many cities. It tells you what your city will feel like in 2080 and gives you a comparison city for context (my adopted Appalachian home will feel like Tuscaloosa), but it doesn't tell you things like "the water table is rising and your basement will never be dry again even though you live 3 hours from the beach").
posted by headnsouth at 8:21 AM on March 23 [1 favorite]

I'm not aware of a decent centralized source for this kind of thing. In part, this is because folks don't seem to agree on how to make the projections. Engineering consulting firms are ready to take giant piles of money to make stab-in-the-dark projections, if you've got more money than time.

For background, I am considering hiring a climate scientist in order to create a set of analyses for a project I'm involved with. We're also probably going to hire a geologist to look at seismic and literal volcano risk, because the usual geotech consulting firms we contacted had no idea how to model things like eruption debris flow or phreatic explosion hazards (non-US project location). I mean, some of them tried, but I could tell they were just looking up sources I already had access to and had reviewed. Ain't gonna pay you to tell me something I already know.

If you're in the US, NOAA has sea level rise predictors that are handy, and the FEMA flood hazard maps are at least a decent starting point.

If you haven't reviewed the IPCC reports yet you should.

For my project we're probably going to design for the upper end of sea level rise projections, design seismic for the maximum credible earthquake, site the structure so that under current topography volcanic flows would go elsewhere, and assume both massively less rain and massively more rain (have both large catchment areas and also large water impound capacity). The US military has a bunch of useful documents on structural durability, by the way.

How to model social/political impacts of any of the above is even more difficult, and given recent geopolitical fuckups by literally every government on Earth we may simply assume that the local government will respond in the worst possible way to any given disaster, but at some point you'll inevitably hit the stage where you realize "well, if all of those things go wrong it won't really matter what happens to our project, so ehhhhhhhh".
posted by aramaic at 8:34 AM on March 23 [1 favorite]

This is one tool that calculates climate risk (and other risks) for a property.
posted by pinochiette at 8:41 AM on March 23 [1 favorite]

For a TLDR-ish forecast, there's this tool. It compares what your city in 30 years will be like compared to a city today.
posted by ocschwar at 9:33 AM on March 23 [1 favorite]

The Climate Explorer is specifically about giving long term predictions based on the best available science.
posted by rockindata at 10:48 AM on March 23 [1 favorite]

The Flood Factor Web Site seems to project out 15 and 30 years, for a specific community or nationwide.
posted by forthright at 5:46 PM on March 23 [1 favorite]

For Canada, there is the climate data website. Also just a note on terminology, these are usually referred to as climate projections. The term forecast is usually reserved for shorter term predictions (ie. from days to seasonal).

The models used have multiple sources of uncertainty related to input conditions, approximation of physics in the models ("model parameterizations") and the inherent uncertainty that comes from modeling a chaotic system. An honest projection of future climate should therefore use multiple models (an "ensemble") run under a range of different CO2 concentrations in the atmosphere ("carbon forcing"), different model specifications and input conditions. The most recent set of models also attempts to represent possible emissions scenarios that would result from different socioeconomic "pathways".
posted by piyushnz at 7:56 PM on March 23 [1 favorite]

It's not my field, but from what I know regional downscaling is one of the areas of ongoing research on the cutting edge of climate modelling. It is much more challenging than producing global predictions. An analogy I have heard is that looking at the energy balance of the whole planet is like comparing how much water is flowing into a bucket versus the amount flowing out, whereas trying to work out where all the energy will be locally is like working out where all the water will be despite chaotic turbulent flows.

Also, one of the biggest sources of uncertainty in all longer-term climate projections is what emissions trajectory the world will follow (as well as the degree to which carbon sinks like forests, permafrost, and methane clathrates end up in the atmosphere). The emissions trajectory is literally unknowable, so there will always be a range of projections. The impact on sinks is probably unknowable too, because there is nothing in the paleoclimatic record as rapid and aggressive as what we are causing now, so there are no comparable cases for comparison.
posted by sindark at 11:08 PM on March 29 [1 favorite]

« Older 2 classical performances by a quartet in one...   |   Don't leave me, Alexaaaaaaaaaa! Newer »

You are not logged in, either login or create an account to post comments