A regional supplier of jet fuel is interested in forecasting its sales. These sales data are shown from the period of 2002Q1 to 2017Q4 (data in billions of gallons):

Date	Sales
Mar-02	23.86
Jun-02	23.97
Sep-02	29.23
Dec-02	24.32
Mar-03	23.89
Jun-03	26.84
Sep-03	29.36
Dec-03	26.3
Mar-04	27.09
Jun-04	29.42
Sep-04	32.43
Dec-04	29.17
Mar-05	28.86
Jun-05	32.1
Sep-05	34.82
Dec-05	30.48
Mar-06	30.87
Jun-06	33.75
Sep-06	35.11
Dec-06	30
Mar-07	29.95
Jun-07	32.63
Sep-07	36.78
Dec-07	32.34
Mar-08	33.63
Jun-08	36.97
Sep-08	39.71
Dec-08	34.96
Mar-09	35.78
Jun-09	38.59
Sep-09	42.96
Dec-09	39.27
Mar-10	40.77
Jun-10	45.31
Sep-10	51.45
Dec-10	45.13
Mar-11	48.13
Jun-11	50.35
Sep-11	56.73
Dec-11	48.83
Mar-12	49.02
Jun-12	50.73
Sep-12	53.74
Dec-12	46.38
Mar-13	46.32
Jun-13	51.65
Sep-13	52.73
Dec-13	47.45
Mar-14	49.01
Jun-14	53.99
Sep-14	55.63
Dec-14	50.04
Mar-15	54.77
Jun-15	56.89
Sep-15	57.82
Dec-15	53.3
Mar-16	54.69
Jun-16	60.88
Sep-16	63.59
Dec-16	59.46
Mar-17	61.59
Jun-17	68.75
Sep-17	71.33
Dec-17	64.88

Please provide relevant graphs and explanations for each part of the answer.

a. Provide an example of a time series graph of these data. What, if any, seasonal patterns are present in the plot? Please explain.

b. Use ForecastX to make a time series decomposition forecast for 2018. Provide an example of a report explaining your forecast. Include a graph of the fitted values, the forecast values, and the actual sales.

c. Develop two other forecasts of jet fuel sales using the following methods: 1. A Winters' exponential smoothing model; and 2. A regression model using just time and quarterly dummy variables. Compare the MAPES for the three models you have developed, and comment on what is good or bad about each of the three models for this application.