Visualize Irregular Water Flow Patterns

Differences in water flow rates for the Colorado River at Lees Ferry, Arizona are instantly recognized in Surfer using a time-series visualization technique developed by Dr. Richard Koehler, CEO of Visual Data Analytics. These patterns practically jump off the screen when viewing one of the temporal raster maps created by Koehler.

The USGS gauge serves as a valuable case study in identifying and analyzing flow. This point is tied to legally agreed upon volumes flow on the Colorado River basin. Lees Ferry is key to monitor the water flowing past this point, a resource vital to the economies of numerous western states.

Water flow is usually plotted on a 2D line graph with time on the X axis and flow rates on the Y. When many years are layered on each other, a ‘spaghetti plot’ is created where details are very difficult to see.

“The irony of line graphs is the more data you display, the less detail you see,” said Koehler, who presented the Lees Ferry case study in a recent Golden Software webinar.

Using Surfer, Koehler shows flow differently. By using “day of year” on the X axis and using “year” on the Y axis, he can represent specific dates by coloring each day. The results show high water flows as blue and low flows as red/orange colors. Suddenly, the magnitude, frequency, duration, timing and change of flows are easy to see. Deciphering patterns in large temporal datasets is challenging; however, with Dr. Koehler’s visualization technique, patterns that are difficult, if not impossible to see on a line plot are clear as day.

1. Glen Canyon Dam regulation – The hydroelectric dam became operational in the early 1960s after which time the flows were significantly altered. Very few extremely high discharge rates, dark blue days, occur after the dam was in place. The flow patterns before and after the dam are striking and easy to see.
2. Spring snow runoff – Until 1963, higher flows from melted snow usually occur in May and June. The beginning and ending of each year’s snowmelt can be seen simply by seeing where days with blue start and end.
3. Drought – The absence of a dark blue line in the midst of the 1930s marks a period of lower snowpack in the Rocky Mountains that left the Colorado River flowing at record-low levels. The red/orange colors of lower flow are easy to locate as these days are surrounded by blue.
4. Sundays and holidays – Discharge from the dam changes with electricity demand. Lower demand means less water moving through the dam. As a result, days such as Sunday or Christmas are consistently represented in orange and yellow colors reflecting the lower flow rates.
5. Individual storms and artificial floods – Isolated blue colors lasting from a few days to many weeks indicate heavy runoff from a major storm. These storms are easily seen among otherwise stable flow regimes. After the dam, artificial floods can be spotted quickly. Comparing the two different types of higher flows is easier with a time-series map.

Deciphering patterns in large temporal datasets is challenging; however, with Dr. Koehler’s visualization technique, patterns practically jump off the screen.

The Golden Software webinar recording featuring Rick Koehler and time series visualization in Surfer is available for viewing here.