3 ms·

Sadly even if we had one sensor per cubic foot of the troposphere nonlinearity would quickly swamp any predictive value, I'm not sure we will ever have accurate

• mturmon 14y ago
  Right, it's generally believed that forecasting will never be effective beyond about 15 days, due to amplification of small uncertainties by the nonlinear atmosphere dynamics. Predictability is better in the tropics by a couple of days, and better in the northern hemisphere winter by a couple of days. It's worth noting that the pressures measured, even if they could be calibrated, would be almost entirely on land, and only at the surface of the Earth, not at higher altitudes (it is, of course, a 3D problem). And also, a lot more than just pressure is needed -- temperature, wind velocity, clouds, aerosols, irradiance, ocean currents, wave height, soil moisture, ...