Start with the axes
Identify the parameter, unit, sampling dates, and station or reach. Check whether the scale begins at zero and whether multiple reaches are being compared.
Look for patterns
Ask whether values are increasing, decreasing, seasonal, clustered after rainfall, or changing only at one station. A trend across several observations is usually more informative than a single peak.
Read the limitations
Missing observations, long gaps, changing stations, and different methods can limit comparisons. Rivistry shows source information and data-quality notes so the graph is not overinterpreted.