Confidence interval very narrow around GAM model

I am trying to plot the trend of my GAM model with a confidence interval around modelled trend, but the confidence interval is very narrow, and I am assuming incorrect. However, I am unsure as to how to correct this.

I uploaded a dataset with data for the daily precipitation and daily increase in streamflow. I named these variables "Precipitation" and "DailyIcrease"

I then created a GAM model comparing the relationship between the daily increase in streamflow and the square root of the daily precipitation depth using the code:

m_pre_sqrt <- gam(DailyIncrease ~ s(sqrt(Precipitation)), data = pre, method = "REML")

Based on the model summary, this model has an r-squared value of 0.885, indicating that the precipitation is a very good predictor of the daily precipitation depth, which logically makes sense.

I plotted the model using the code:

plot(m_pre_sqrt, residuals = TRUE, pch = 20, xlab = "Square Root of Daily Precipitation Depth (mm)", ylab = "Square Root of Increase in Daily Stream Discharge (m3/s)", family = "A", xlim = c(0,8), ylim = c(0,12))

This code returned the following plot.

As shown by the plot above, there is a very narrow confidence interval, which leads me to believe that something is wrong. As shown in the plot above there are quite a few points between 2 and 5 mm that deviate from the regression curve, however, the confidence interval does not reflect this. I am thinking that something is definitely wrong. Does anyone have any advice? TIA

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.