This whole article boils down to one thing. We know that the earth is warming and is going to continue warming but we can’t model the effects on a local scale. The problem is simply too complex for our current models and our current computing infrastructure.

The absolute best climate models have horizontal resolutions on the order of 10km. That means that there is one value for temperature every 10km, one value for water vapor every 10km, one value for north-south wind speed every 10km, one value for east-west wind speed every 10km.

Of course we can’t model hyper local climate effects! We have one pixel every 10km to help us understand an immensely complex system and how it is changing over hundreds of years!

Hell, we can barely model hyperlocal weather effects! We have high resolution satellite measurements every 10 minutes over the entire globe. We know where everything is and can give the models very good initial conditions. Yet we still can’t say whether it will snow in your neighborhood tomorrow. We can give a probability, but nothing certain.

If, given high resolution initial conditions for modeling weather, we can’t model a neighborhood’s weather accurately, why would anyone think we can accurately model hyperlocal climate effects?

Source: A drunk atmospheric scientist