Pandas: Impute a given number of missing values before/after a series of available values

Question

Let\'s say I have a time series where I usually have data available for a certain continous span of years, but missing values before and after that span, like this: