4 ms·

I believe the hurricane tracks is a solution to the problem outlined in the HN post 3 days ago on "Misinterpreting hurricane forecasts"[0]. The hurricane track

by duanem 7y ago

I believe the hurricane tracks is a solution to the problem outlined in the HN post 3 days ago on "Misinterpreting hurricane forecasts"[0]. The hurricane tracks are a new addition to Flowx but I believe it can be improved a lot. For example, you can color the dots by the hurricane category or wind speed. Better still, draw a heat map around each dot would generate a probability map for the hurricane. [0] https://news.ycombinator.com/item?id=20833030 https://news.ycombinator.com/item?id=20833030

astrophysician 7y ago
Not a bad idea but only if the tracks are independent. Was reading something saying that these tracks aren’t different random realizations of the same simulation (in which case what you’re saying is a good way to estimate the posterior of the hurricane track — it’s just Monte Carlo) but that these tracks come from many different types of simulations, some with higher quality than others and they’re likely all correlated in some weird way. If it’s the latter case there has to be some care and thought put into combining the predictions into a single estimate of the posterior.
- duanem 7y ago
  The initial conditions of the weather simulation, which is based on measured data and previous simulation results, are perturbed to within a reasonable range to represent the error in the measurements. Because of the equations being solved and numerical methods being used, these small changes can cause large change on long term predictions. The tracks come from one original simulation. If you look at the start of the tracks closely, you'll see the eye of the hurricanes diverge in a regular grid, so I'm guessing they are perturbing the direction and speed of the bulk movement of mass - I would have to research more. So the initial conditions and early stages of the simulations are correlated, but each simulation is solved independently. The question is how long into the future can you consider each track less-correlated. Given the met organizations release probability envelopes based on these track means it's a reasonable approach. What I really want to do is weight the predicted wind speed with the probability map. I don't know if this will work. But you can imagine if a hurricane has a 75% change of going well into the continent it would weaken and cause less damage, but if the other 25% of the tracks go over sea and hit an island, it'll likely strengthen and cause more damage. In any case, all the above will be interesting.

by duanem 7y ago

astrophysician 7y ago

Not a bad idea but only if the tracks are independent. Was reading something saying that these tracks aren’t different random realizations of the same simulation (in which case what you’re saying is a good way to estimate the posterior of the hurricane track — it’s just Monte Carlo) but that these tracks come from many different types of simulations, some with higher quality than others and they’re likely all correlated in some weird way. If it’s the latter case there has to be some care and thought put into combining the predictions into a single estimate of the posterior.

duanem 7y ago
The initial conditions of the weather simulation, which is based on measured data and previous simulation results, are perturbed to within a reasonable range to represent the error in the measurements. Because of the equations being solved and numerical methods being used, these small changes can cause large change on long term predictions. The tracks come from one original simulation. If you look at the start of the tracks closely, you'll see the eye of the hurricanes diverge in a regular grid, so I'm guessing they are perturbing the direction and speed of the bulk movement of mass - I would have to research more. So the initial conditions and early stages of the simulations are correlated, but each simulation is solved independently. The question is how long into the future can you consider each track less-correlated. Given the met organizations release probability envelopes based on these track means it's a reasonable approach. What I really want to do is weight the predicted wind speed with the probability map. I don't know if this will work. But you can imagine if a hurricane has a 75% change of going well into the continent it would weaken and cause less damage, but if the other 25% of the tracks go over sea and hit an island, it'll likely strengthen and cause more damage. In any case, all the above will be interesting.