Search Our Site
Trends: During October, streamflow peaks and low points were coming down from the significantly high levels at the beginning of the month caused by tropical moisture tracking our way. By Nov 3 (before up to 4 inches rain), 20 out of 63 stream gages had dropped to flow rates near normal for the day of year. 37 were at still at high flow levels, 6 were still very high compared to their records for the date. This appears to be similar to the pattern seen in August.
The 2 to 4 inches of rain that fell over the northwest and north central parts of the state Nov 3-4 dramatically broke the pattern of decreasing peak flows. The northwest half of the state was in a path of heavy rain that started about 50 miles across in western Pennsylvania on Nov 2 then widened as it traveled northeast until it covered most of Maine. Stream gages with watersheds in that path showed sudden jumps to levels equal or higher than any peaks since August 1. Although there were 17 reports of heavy rain to the National Weather Service, there were no reports of flooding. Click here for some news stories.
Bunnell (Burlington) Brook is a good example:
Coastal CT got an inch or less from the storms of Nov 3-4 so the stream gages there do not show much change from their previous flow patterns.
Storms began consistently tracking over Connecticut in early September (see graph below), resulting in streamflow peaking at higher and higher levels and coming often enough that the flow rates between storms also became progressively higher. An analysis on October 3 showed that 11 of the 77 USGS gaging stations in or near Connecticut reported flow rates that exceeded the two-year recurrence level (a level statistically not expected to occur more than once every two years) for those stations. Six of those also exceeded their 10-year occurrence level.
The result of new rainfall on saturated soil running off to streams still high from previous storms was that more flooding occurred with every storm. From Sep 12 to Oct 3 there were 64 reports of flooding to the National Weather Service and many news stories. There is some general information about flooding below. Remember: Turn Around, Don't Drown!
From Aug 1 to early September, a pattern similar to October was seen. The graph below is an example of Connecticut streamflow patterns since the beginning of August that shows periods of unusually heavy rainfall separated by dry weeks. The first three weeks of August saw high streamflow during storms as expected, but in most cases streamflow did not drop back to normal before the next storm hit. Then, during the last week of August and the first week of September, storms were diverted around us by a persistent high pressure system, causing flows to fall steadily until most stream gages were reporting near normal flow rates for the time of year.
Click the graph below to compare all CT stream gage graphs for Aug 1 - Nov 5 (very large download, takes a while to load).
We define high flow as at or over the stream's 75th% percentile for that date, very high at 90% or higher, from the data file at https://waterwatch.usgs.gov/webservices/realtime?region=ct&format=xml when rivers are near their low points between storms.
Why do we count? The flow between storms is an indicator of long-term stream conditions. Although expansion of a river or stream channel onto its floodplain is a normal, natural occurrence, human interference can cause higher than normal and more frequent flooding. The higher the water level remaining in streams and rivers before a storm, the more likely there will be dangerous conditions, streambed scouring, and other detrimental effects of high water.
When water levels are low, portions of the streambed become too dry to support the aquatic life that usually colonizes those areas. Groundwater seeping from the stream banks can help organisms survive for short times; therefore, well measurements can also be used to identify areas where aquatic life might be in jeopardy.
A count of the number of USGS stream gages reporting low flow at their lowest point between storms is valuable as an indicator of conditions statewide. If the number of streams that reach low levels between storms is increasing over time, it raises concern for the recovery of stream ecosystems from dry conditions of previous years. If the count decreases, we can look for recovery. If most of the low-flow gages are in the same area of the state, it could indicate a regional problem. We also note any record low flows for the date (see Sep 14, 2017 discussion).
Brief high water, even flooding, caused by sporadic heavy rain does little to alleviate long-term effects of dry conditions, because it takes time for ecosystems to recolonize streambeds after a drought has killed off aquatic organisms. Direct observation of any particular segment of stream is necessary to determine actual streambed conditions, because there are only 62 stream gages in Connecticut and the conditions they report may not represent the conditions in the hundreds of streams with no gages. A discussion of normal streamflow patterns can be found here.
In Connecticut, the USGS has 10 real-time groundwater monitoring wells. All of them show large rises in the water table over the past months to well above normal for the time of year. The rise began in August for two them; the rest showed more modest increases until September.
From the end of September until the major rain events of Nov 3-4, the water level in those wells had shown small increases and decreases as storms added groundwater, but their levels stayed pretty steady or had fallen slightly. Groundwater levels rarely show much change due to a few minutes of heavy rain, but on Nov 3-4, all the wells showed a jump due to the long duration of the rainfall that caused the maximum rate of infiltration to occur over many hours.
On November 5th, of the 6 real-time well reports that compare the current depth to water table with the average for the date over multiple years of record, 5 were showing high water tables. Two of those were breaking high groundwater records for the date over 7 and 8 years of records. Only 1 well of the 6, in Greenwich, was near normal.
The USGS Groundwater Watch website displays data from 60 other wells that are measured once a month, comparing the most recent measurements to each well's measurements taken in that month in previous years. With October rainfall closer to normal than September, well levels became more normal for the time of year. At the end of October:
12 of the 60 wells were near normal (there were 7 in September)
Monitoring the water table can provide clues to future conditions. There was an apparent pattern from 2013 to 2016 in which normal groundwater levels in winter and spring alternated with increasingly well-below-average levels in summer and fall. Although 2017 groundwater levels dropped as far below average as in 2014 or 2015, they did not drop as far as in 2016. This year began to look like 201, but the water tables began rising again in mid-July to mid-August so that this year's lowest water table levels were higher than any low point in at least five years. Click here to see all the CT USGS real-time groundwater measurements that compare levels from the last four years to the median level for each day of the year (lots of data, so it takes a long time to load).
Groundwater levels influence streamflow. The water level in streams and rivers between storms depends on the flow of water out of the ground into the stream channel. The higher the water table, the more water is available to keep streams flowing. During a drought, smaller upland streams dry up faster than the larger rivers (click here for details from a previous Know Your Flow! webpage). An occasional brief, powerful rainstorm does not do much to help groundwater levels rise. The rain falls faster than the ground can absorb, and thus much of it becomes stormwater runoff. In contrast, public water supply systems that depend on reservoirs benefit from heavy rains because the reservoirs collect the runoff. The runoff, however, can wash pollutants into the reservoirs, lowering water quality.
High water tables keep streamflow from dropping very far or very fast between storms, so when the next storm hits, the runoff pushes streams to higher levels with more flooding.
Total October rainfall varied widely across the state. Southeast and southwest CT received from 25% to 100% above normal, while small areas near New Milford and New Hartford got 50% of normal for the month. Including the recent heavy rain and the wet weeks of September however puts all of the state well above normal for the 60 days before Nov 5 . Small areas of Bloomfield, Windsor and East Windsor received over 600% of normal in the past 60 days. Go to the National Weather Service (NWS) Advanced Hydrologic Prediction Service at this link for updated, detailed maps and data about precipitation amounts for Connecticut at various time frames. Normal November rainfall across Connecticut ranges from 3 to 5 inches for the month, so a month of rain fell on November 3-4 in half of the state. On Nov 5, one-week forecasts were predicting another inch or two, then a long-range (15-day) model was showing storms mostly missing us for the week after.
Click here for a prediction of total precipitation for the next 15 days.
For technical information about past broad-scale weather patterns, see NOAA's National Centers for Environmental Information State of the Climate Synoptic Discussions.
Your water utility and town are the best sources of information on the status of your water supply. A few systems post their current capacities regularly on their websites.
Aquarion maintains on its website a weekly updated water usage report for their systems in southwest CT. The report includes water demand graphs, system transfers, system capacities, days usable storage remaining, and precipitation. As of Oct 8, all those systems were well above where each system usually is at this time year based on a 20-year average. As of Oct 1, MDC reservoirs were at "99% of capacity with over 39 billion gallons in storage representing 657 days of supply assuming typical water use and no precipitation." The South Central Regional Water Authority reported that on Sep 30, its reservoirs ranged from 82% to 100% of capacity, averaging 86% of capacity, which is 18% higher than usual for the time of year.
The most recent CT Department of Public Health's Monthly Reservoir Status Summary was dated July 13. At a Water Planning Council meeting, a DPH representative reported that a new, more comprehensive and timely report system should be online in January.