Geomagnetic Field Disturbance Predictor (ML Model)

This was our final model which can measure the disturbances in the interplanetary magnetic field IMF. Geomagnetic storms are caused by the interaction of solar wind with Earth's magnetic field. The resulting disturbances in the geomagnetic field can wreak havoc on GPS (Global Positioning System) systems, satellite communication, electric power transmission, and more. These disturbances are measured by the Disturbance Storm-Time Index, or Dst.

The primary input data is provided by sensor data from two satellites, NASA's ACE, and NOAA's DSCOVR. The problem we faced while handling was:

Data distributed in three different files.

The units of the timestep were different in three files the plasma readings where in terms of minutes the disturbance storm index was in hours and the sunspot numbers are on monthly basis.

There was a lot of missing values

We padded (forward fill) the data so that every data is in terms of same timestep and we followed interpolation to fill the missing values.

This model used the instantaneous data from the satellites this data can be used to give hourly forecasts. Prediction of Disturbance Storm Index has a variety of applications which include calculating the risk related with a particular flare, Predicting the Auroras and gives us a higher level of understanding on geo magnetic field