Many Texas cities have experienced substantial growth in population over the last 20 years. The growth of Houston, Texas, since 2015 is shown in the scatterplot. A least-squares equation that relates the number of years since 2015 to the population of that year (in millions) is given by Population hat = 0.025 (year) + 2.01. A graph titled Houston Population versus years since 2005 has years since 2005 on the x-axis, and population (millions) on the y-axis. A line of best fit goes through (4, 2.1) and (10, 2.25)

A graph titled Residuals versus years since 2005 has years since 2005 on the x-axis, and residual on the y-axis. The points curve down, and then curve back up.

Based on the residual plot, is a linear model suitable for modeling the population growth of Houston?

A linear model is suitable because a pattern is evident in the residual plot.
A linear model is not suitable because the residuals do not cluster around zero.
A linear model is not suitable because the residual plot shows a curved pattern.
A linear model is suitable because the residuals seem to be randomly scattered.