Earth Observation data (including satellite observations in the optical as well as the microwave domain, complemented by in-situ measurements) provide us the unique possibility for estimating the states and dynamics of geophysical parameters on a global scale. These include meteorological parameters (e.g. temperature), vegetation parameters (e.g. vegetation optical depth), hydrological parameters (e.g. soil moisture), and many others. For retrieving the state of such parameters from measurements of electromagnetic waves it is necessary to understand and model the geophysical processes responsible for the interaction between the observed electromagnetic waves and the parameter under observation. However, the complexity of geophysical processes requires us to make various assumptions, simplifications and approximations in such model in order to be able to convert our measurements into parameter estimates. In this lecture you will gain an understanding of how such model can be created, why and how assumptions and simplifications enable us to invert the model, and how to get estimates of the uncertainties of the retrieved parameter estimates.