Blowing away the fog of ignorance.
Oldweather is going great guns (literally in some cases),
three six ships completed already, and we’ve started to generate scientific results. So maybe this is a good time to step back and remind ourselves what we hope to get out of it.
One of our main aims, for the weather observations recovered by the project, is to improve our reconstructions of past weather conditions, and we want to reconstruct not just the surface pressure, sea and air temperatures, wind speed and direction, that we are digitising from the logs, but the complete state of the atmosphere. This will give us air temperatures and movement (both at the surface and higher up), clouds, rainfall, humidity … Remarkably, we can do this – meteorologists have built sophisticated computer models of the atmosphere, and we can use these models to calculate the complete state of the atmosphere just from our limited set of surface measurements. This is not easy to do, but we’ve a lot of experience doing it, because this is how weather forecasts are made: collect observations of present weather, use a computer model to synthesise those observations into a representation of the state of the whole atmosphere, run the models forward from that state to see what will happen tomorrow.
So we’ve teamed up with colleagues in the US who are running the Twentieth Century Reanalysis Project, they are essentially running a weather forecast not just for today, but for every day in the past 140 years. Their results aren’t as good as the reconstructions of present-day weather (present-day reconstructions can use all the modern satellite observations), but it does work. Well … It works pretty well … in many times and places … where there are enough observations.
The figure shows such a reconstruction, for March 8th 1918. Actually it only shows sea-level pressure and surface wind, but we can make similar figures for 500hPa height, tropopause temperature, jet streams, any atmospheric parameter. The coloured circles mark observations from ships and weather stations, the coloured contour lines and black arrows show the reconstructed pressure and wind speed.
There’s one more feature in the figure, however, and that’s the fog. It’s not real fog, but a metaphor. I’ve added fog to the figure in places where the reconstruction is very uncertain. It’s a marker for ignorance, and the ignorance is the result of not having any observations. One observation in the right place can be enough to make a good reconstruction over a considerable range, so the million or so observations Oldweather will produce can make an important difference. We can blow away this fog, and reveal the weather hiding behind it.