(please sent comments or questions to Michael W. Finkbeiner, LS email: firstname.lastname@example.org
Area above USGS gage station at Comly Ave= 25.6 sq.miles = 16,384 acres.
Large volumes of water are measured in acre-feet. One acre of surface 1 ft deep. That's a 300 x 160 ft football field 10.9 inches deep.
Engineers rate storms by their return frequency for a given time of storm ( 1,2,6,12 or 24 hours) and
frequency of years ( once in 1,2,5,10,25,50 and 100 years, which is an annual chance of 100%, 50%, 20%, 10%, 4%, 2% or 1% respectively)
The typical maximum storm for design purposes is the 24-hour 25-year storm. That is engineer-speak for the worst expected day of rain with a 4% annual chance of occurrence.)
By empirical record keeping, that storm according to the local Flood Board is 5.75 inches, but estimates vary from 5.3 inches to 6.3 inches:
25-year 24hr storm = 5.75 inches of rain in watershed acres=0.48 ft or 7,850 acre-ft of total water.
A two-inch rain at 0.2 inches/hour raises the discharge rate from 100 cfs to 1000 cfs in 5 hours.
At 1000 CUBIC FT/SECOND, the river discharges 3600 sec/hr*1000 cf/sec / 43560 sqft.per acre= 82.64 ACRE-FEET PER HOUR
At 2000 CuFt/sec, the Byram River bridge at Comly Ave is passing an acre-ft of water in 21.78 seconds.
The objective of watershed management and storm water control is to shave off the peaks in the graph below, and moderate the damaging discharge. This is accomplished by either early release from controllable structures at dams and berms, which makes the ramp up to the peak less steep, or detaining stormwater until the peak passes.
A typical detention area wetland might be 2 acres. Detaining four feet of stormwater holds 8 acre ft. or 6 minutes of river discharge at the 1000 cfs rate; at the more typical 100 cfs, the area could detain 60 minutes of discharge.
|Record of 10 storm events, two of them back-to-back storms, showing the increase in discharge caused by reduced storage. Most engineering calcs assume a dry start for the designed storage, with gradual leakoff. This is only due for single-day storms with a dry period to follow.|
|The US Geological Survey has installed a watershed monitor station at the Comly Ave. Bridge, with an on-line rain gauge and a river gage-staff (their spelling), which uploads data to http://waterdata.usgs.gov/nwis/uv?site_no=01212500 at 15 minute intervals. This permits a near real-time comparison of water volumes falling into the 25.6 sq.mile area upstream, with the actual discharge of run-off to Long Island Sound.|
| The Storms of March 29 and 30 represented
minor events individually, yet in combination they acted with flooding
consequences far beyond expectations of their actual intensity and
volumes. See analysis below:
Text of Pre-Event MEMO from EarthImage 3/29/2010
to Greenwich Flood Board (http://www.greenwichct.org/FloodErosion/FloodErosion.asp)
Two Points of Immediate Concern for the Byram River:
During the April 15, 2007 storm event, the dam at Glenville had water released with what many perceived to be unfortunate timing for the Pemberwick neighborhood downstream.
While I understand the Flood Board is not the same department or agency as Emergency Management, nevertheless some advisory coordination is in order. With the full moon high tide predicted for 11:30 am today at 8.4 feet over Mean Low Water and the storm surge already approaching two feet over that, any release of water from controllable spillways should be done mid afternoon today, well before conditions worsen. This would apply to Putnam Lake as well as the Bargh, if Aquarion participates in floodway management, as they should.
During a tidal flood event a few months ago, I observed on the north side of the Rt. 1 Byram River bridge a strong north-bound upstream current of salt water intrusion, with fresh storm water flowing south on top and over the heavier salt water below. So the tidal cycle is a factor north of Rt. 1 on the Byram. A river discharge of 6500 cfs at high tide may submerge the bridge, especially if tree debris from the prior storm is picked up and caught at the bridge arch.
On-line near real-time tide data with storm surge mapped is available at http://tidesandcurrents.noaa.gov/data_menu.shtml?stn=8516945%20Kings%20Point,%20NY&type=Tide%20Data
River Stage Gage Station:
Now that our Byram River gage station is up and running, we can empirically understand of actual operation of the watershed much better. It has been my contention that the engineering design assumptions and parameters, which are implicit in the approved drainage structures throughout small urban watersheds, fail to work as planned during storm emergencies.
Specifically, I am concerned that the roughness co-efficients, natural storage, run-off co-efficients, and travel velocities all worsen significantly on the second and third days of multi-day storm events. As you know, most engineering designs and reviews are based on a single 24-hour storm event, and not a multi-day storm cycle.
The river stage, discharge hydrograph, and rainfall amounts in 15 minute increments now give us the means to study this process empirically. Data with graphic presentations is available with an approximate two-hour delay at :
Post Event Analysis of Watershed Response (revised 3/31/2010):
worst of the heavy rain was endured by southern New England, where rain amounts
exceeded a half a foot in some locations. Through Tuesday evening, North
Kingstown, R.I., topped the rainfall totals list with 9.47 inches.
rainfall amounts from Sunday night to Tuesday evening are given below:
Providence, R.I.: 8.75 inches
R.I.: 8.74 inches
River, Conn.: 8.44 inches
Hollow, N.Y.: 8.03 inches
Mass.: 7.12 inches
Conn.: 7.02 inches
N.Y.: 6.89 inches
Mass.: 5.59 inches
rain pushed this month's total to 16.32 inches (through Tuesday) in Providence,
R.I. This March is now the city's all-time wettest month on record.
is experiencing its second all-time wettest month with 14.83 inches measured
this March (through Tuesday).
The following is analysis of the Byram River Watershed Rainfall input
less run-off out at Comly Ave Gage Station
for the week of the storm event March 29-30, 2010:
Rainfall began at 21:30 hours on 3/28/2010 staying under 1/2 inch per hour intensity for a total of 2.5 inches of rain.
First day's net input to watershed peaked at 2330 acre ft at 14:15 hours on 3/29.
The second day's rainfall continued at the same peak intensity, accumulating another 3 inches.
The second day's input to the watershed peaked at 4465 acre feet at 11:30 hours on 3/30.
Thereafter, net detention of the peak volume fell at the following rates, until it all ran downstream, leaving net zero.
90 % remained in the watershed for 0.125 days (3 hours)
80 % remained in the watershed for 0.26 days (6 hours)
70 % remained in the watershed for 0.41 days
60 % remained in the watershed for 0.635 days
50 % remained in the watershed for 0.948 days
40 % remained in the watershed for 1.438 days
30 % remained in the watershed for 2.12 days
20 % remained in the watershed for 3.063 days
10 % remained in the watershed for 4.292 days
0 % remained in the watershed after 5.91 days
Thus we see 50% of the accumulated rainfall in the watershed was discharged in under 24 hours after the peak storage time,
and the remaining 50% was lost to run-off in less than 6 days.
These are numerical facts from the government data set.
Where is the natural detention and absorption?
Binney Park at Perrot Library
Bruce Park at skating pond
Sundial at Bruce Park
|This graph of discharge rate per 15 minutes shows
the significant difference of the second day, when storage and roughness
of the dry conditions in the watershed had been eliminated, leading to a
more severe peak rate of 200 acre ft per hour. That is the amount
of water held in 200 acres of wetlands 1 ft deep, or more realistically,
since natural wetlands are not modified to detain a one-ft depth, it is
the same as 2400 acres giving up an inch of impounded water in an hour.
|The cumulative discharge shows the total amount of run-off from the storm event.|
Tree damage and wind-throw in the forests during the March 14 Storm. (Forester's Version)
Driving around Greenwich would lead one to think that major tree damage was wreaked upon the whole Town. Yet take walk in any of our woodlands and you will find very little evidence of storm damage. Why is this? Is God testing homeowners and drivers, but not woodland animals?
One explanation is that in the forest there is intense competition for light and nutrients, and the soils are relatively undisturbed. However, in Town roadways there is extra light and nutrient run-off from homeowner fertilizer use. Correspondingly, the tree root systems are cut, filled upon and the soil compacted.
In the 25 years since Hurricane Gloria, these trees have accumulated bigger and bigger topsails, while supporting this wind load with poor and poorer root systems.
Thus the damage is localized to homeowners, and especially to the power and telecom overhead wires along the highways. CL& P's efforts to prune their lines have been consistently opposed by the Tree Warden and the Town's people.
Thus is it a surprise that the CL & P efforts looked like this a week after the storm:
From March 21, 2010
CL&P power outage report @ 3 pm
Accumulated Data History of Input and Output at the Gage Station (the watershed water budget.)
Total History of Gage Station shows the water budget for the system. This is the running sum of rainfall input. Twelve inches of rain accumulation at the gage makes for 1 foot of water over 16,384 acres, or 16,380 acre/ft. (1 acre-ft is a football field 300 x 160 ft 10.89 inches deep in water. This is 325,872 gallons.)
The gage station also report cubic feet per second discharge of the Byram River in 15 minute intervals. The running sum of these numbers is the total liquid drainage from the watershed. The difference is water left behind, which divides itself into what soaks into the ground (aquifer recharge for well and ground water), what is taken up by plants, and what evaporates.
USGS reports that this amount is typically 50% of storm water input, the other half being run-off. However, the Byram watershed has two wildcard characteristics: poor soils compared to the deeper gravels of the Mianus system, and a diversion tunnel to Putnam Lake, serving the reservoir interests of the local water company, and adding some discharge to Horseneck Brook, which flows along Lake Ave.
The Byram River Gage station also allows us to track total rainfall into the watershed against run-off to Long Island Sound. The last 120 days look like this:
Units on the left-vertical axis are total acre-feet accumulated in the 16,384 acre watershed. Horizontal axis is 15-minutes time units for 120 days. When the red line climbs to 16,384 one foot of water will have been captured as non-run-off. This is water we can use. The rest sleeps with the fishes.
During the winter months, many storms produce 100% run-off. If snow from prior storms is melted by new rain, the run-off can exceed 100%.