We analyse methods for detecting change points in optical remote sensing lake drainage time series. Change points are points in a data set where the statistical properties of the data change. The data that we look at represent drained lakes in the Arctic hemisphere. It is generally noisy, with observations missing due to di!cult weather conditions. We evaluate a partitioning algorithm, with five di↵erent approaches to model the data, based on least-squares regression and an assumption of normally distributed measurement errors. We also evaluate two computer programs called DBEST and TIMESAT and a MATLAB function called findchangepts(). We find that TIMESAT, DBEST and the MATLAB function are not relevant for our purposes. We also find that the partitioning algorithm that models the data as normally distributed around a piecewise constant function, is best suited for finding change points in our data.