a. A time series is a collection of data points measured at successive points in time, typically ordered chronologically. Analyzing a time series involves identifying underlying trends, seasonal patterns, and cyclic movements to make predictions. The analysis can be done through smoothing techniques like moving averages or using statistical models such as regression analysis. The purpose of analyzing time series is often to forecast future data points by continuing the identified trends into the future.

b. The two models used to estimate seasonal variation are:

1. Additive Model: This model assumes that the seasonal variation adds a fixed amount to the trend in each time period. Variations above the trend are positive, while those below the trend are negative. To calculate, the actual historical values are compared to the trend, and the differences are grouped by time period to find the average seasonal effect.
2. Proportional (Multiplicative) Model: In this model, seasonal variation is expressed as a percentage or proportion of the trend. The actual values are divided by the trend values to find seasonal factors, which are applied to adjust future forecasts based on the identified seasonal effect.

Explanation of each model includes its respective method of accounting for seasonal variations in data points.