graph_weather
graph_weather copied to clipboard
PyTorch implementation of Ryan Keisler's 2022 "Forecasting Global Weather with Graph Neural Networks" paper (https://arxiv.org/abs/2202.07575)
I'm currently running a residential solar system and using online forecasting predictions to set battery charge levels as to not over charge during off peak times. My current provider has...
https://arxiv.org/abs/2203.12297 ## Detailed Description The paper shows making a Gan that can up sample global forecasts for precipitation to high resolution local forecasts, which is similar to something we want...
## Detailed Description Courtesy @JackKelly but satellite imagery seems very useful for predicting PV output, and to go out more than a few hours, such as day ahead, etc. the...
https://www.maelstrom-eurohpc.eu/products-ml-apps.php#dataset-for-energy-production-forecast contains seemingly quite a lot of data with ensembles already available that might be useful to try out.
## Detailed Description The paper uses land/sea mask, and orography masks with 0.1 degreex0.1 degree resolution for extra input. Currently, these inputs have to be added to the same nodes...
## Detailed Description GFS data is publicly accessible and so we can make that dataset open, as well as its one of the ones used in the original paper, although...
## Detailed Description If the model is fast enough at inference, it could be quite easy and cheap to generate a lot of forecasts at once. The original implementation took...
## Detailed Description Instead of just the next state, we could condition the model on the lead time (1 hour, 3 hours, 6 hours, 12 hours, etc.) like in MetNet-2....
## Detailed Description For adaptive meshes in #3 we need ground truth at various resolutions. One possible way would be to train on the world with the GFS 0.25 degree...
## Detailed Description Being able to get more detailed output around areas of interest is quite important, such as around the UK or some other country of interest. ## Context...