The Atmosphere As A Sponge

Think of a sponge. If you pour water into it the sponge will absorb it… until it can’t.

We’re surrounded by this colorless, odorless, transparent thing: our atmosphere. It’s full of many components suspended in different ratios depending on where or when you measure.

With me so far? This won’t take long.

Depending on temperature the atmosphere has a finite limit of how much water vapor it can hold. Too much and it becomes saturated. As long as the atmosphere is below saturation (aka 100% humidity), most of the water vapor entering it will be held.

Tonight it’s snowing like crazy ‘over’ all of Connecticut. At the same time it’s only the shoreline where we’ve reached saturation.

At Bradley that overhead snow is evaporating… disappearing as water vapor into the clouds… and hasn’t yet reached the ground.

Think of a sponge. If you pour water into it the sponge will absorb it… until it can’t. That’s saturation. After that everything you pour in will pass right through!

On the shoreline this manifested itself by the very rapid transition from light to heavy snow.

2 thoughts on “The Atmosphere As A Sponge”

  1. Well – the atmosphere is finally wringing itself out, Geoff. The snow just began in Manchester within the last hour – kind of fine snow – but coming down moderately.

  2. Current numerical models rely on finite expressions of continuous equations based upon conservation of mass, motion, water, heat/energy, and gasses. These models seem to be based on micro level physics specifications, rather than on ongoing macro level observations. These ‘physics” models may work well at explaining the “why” of weather phenomenon, but aren’t designed to be predictive. They are hypothesis testing models, designed to “prove” rather than forecast. In scaling these models to the meso level, they rely upon extensive “forcing” assumptions and imposed conditions and constraints. Using these models for prediction results in aggregated micro effects (including associated errors) being scaled to meso level by relying upon an infinitely scalable model viewpoint. They do provide useful data on elasticities.

    The discontinuity between these models and your “gut” reading of empirical data (however represented in graphics, tables etc.) causes your “angst” over forecast accuracy. And you are right. Your extensive experience in empirical observation tells you something different (and better, and more relevant) than the physics models do.

    I think the goal of local meteorologists should be to deliver a prediction that focuses on temperature, precipitation and wind speed on a short temporal basis (say 24 hours) and a fine spatial basis (say 5 km). That would be useful! But to do that, you would have to build your own prediction model – probably using convergent simultaneous linear equations, and forget the fancy, sexy physics models. Sometimes simpler is better.

    It’s interesting to me that the AMS has no award for best forecaster, but lots of awards for research.

Leave a Reply

Your email address will not be published. Required fields are marked *