Providing non-linear satellite state models

A computer system comprising:

• one or more hardware processors;
• at least one memory coupled to the one or more hardware processors and storing one or more instructions which, when executed by the one or more hardware processors, cause the one or more hardware processors to:
• obtain first global navigation satellite systems (GNSS) signal data from a set of GNSS satellites;
• obtain first satellite state data comprising estimated orbit data valid at one or more particular times for the set of GNSS satellites;
• generate a non-linear satellite state model comprising a plurality of model parameters by adjusting the plurality of model parameters using one or more machine learning techniques based on the first GNSS signal data and the first satellite state data, wherein the non-linear satellite state model generates output satellite state data based on GNSS signal data;
• and
• generate GNSS satellite state analysis instructions configured to allow a second computer system to access the non-linear satellite state model to generate updated satellite state data for one or more particular GNSS satellites by applying the non-linear satellite state model to particular GNSS signal data corresponding to the one or more particular GNSS satellites.