Many water quality factors depend on land uses, flow rates, interchange with groundwater or hyporheic flow, erosion, sediment, temperature, etc. Some things may lend themselves to trend analysis, but many not. Biological indicators such as macroinvertabrates have some memory in that they respond to pollutants, toxins, loss in habitat, etc. The researchgate paper ~1975 on Spring Water Quality in Missouri showed some water quality trends with discharge in karst terrain. It makes sense that soluble rock would exhibit some trend with residence time as higher discharges from rainfall dilute some factors and have less residence time for soluble reactions, yet also flush others from surface runoff. If you have sampled enough to establish a long term trend, then you will want to establish explaination, and also any conditions where trend is reversed or altered, explain any substantial outliers of data, why? Trends may show up when plotting data, or in multivariate analysis with high correlation of factors.
This depends on the time interval of the data. For example, yearly data has no seasonality, monthly data has a seasonality of 12 months......etc. This depends on the sampling time interval as we illustrated. Now if you have time series analysis for a water quality data then please inform me what type of water quality data is that? so that I can help you dear collageous. Good luck, and I completely agree with the opinion of Mr. William F. Hansen. He is an expert.
There are some simple non-parametric tests for trend that take into account seasonality of the data. You might look at the Mann-Kendall test: http://rcompanion.org/handbook/F_13.html .
- Badges
- Science topic