I haven’t written anything to date for this analysis. I’ve simply manipulated data from the USGS Suspended Sediment Database and summarized them in a spreadsheet available from the Ohio DNR and made a map. The watersheds that are delineated in the map were made using EPA Watershed Delineation add-on in HydroDesktop. The one data point that jumps out at me is the 1982 annual load for the Salmon River: 1,158 tons/mi2/yr, 96% of which occurred on June 5th and 6th. (Granted, the record that year only had 245 days of data, and no data were collected on June 4th). In any event, the number is still quite large and puts it in league with some of the larger annual loads I saw while working in coastal northern California, an area known for its high sediment loads. The subsequent question that makes me curious is how did the channel respond to this flood in early June, 1982? I have to assume that the massive load of sediment was deposited somewhere, but I also have to assume that much of bed and banks were scoured. The Google Earth imagery goes back to 1990 and it suggests a fairly stable single thread channel that hasn’t migrated much in the past two decades. The vast majority of the floodplain appears forested, so my assumption is that much of the suspended sediment was deposited on the floodplain on the receding limb of the storm event or it was simply carried downstream to the Connecticut River.
(Circling back to the Allen Brook analysis, these results from CT gauges make me question all the more the estimated sediment loads using the SWAT model for subbasins 1, 3, 4 an d 22 which supposedly were generating annual loads in excess of 5,ooo tons per square mile. As consulting colleague of mine in California once said to me, “Beware the uncalibrated model.”)
The other bit of insight I gleaned from these data are how episodic the pulses of suspended sediment are in New England. Granted, the data are limited: only 3 gauges had 5 or more years of nearly complete (360+ days of data for a given water year). Nevertheless, it still suggests to me that most rivers are relatively calm and not transporting an inordinate amount of suspended sediment and then WHAM!, a large and rare event occurs that has the ability to deliver a substantial amount of sediment, well outside the range of normal annual load variability (very roughly speaking 20 to 100 tons/mi2/yr).
I also used the Effective Discharge from Suspended Sediment spreadsheet available from the Ohio DNR here. Unfortunately, I’ve never quite gotten the gist of the spreadsheet, despite looking at it and reviewing the hidden Calculation Table tab. In any event, I’ve plugged in the data for the eight CT gauges. The hyperlink in the first column should take you to the USGS data, and the hyperlink on the right should take you to the spreadsheet I’ve saved on Box.net. Each spreadsheet is just under 6Mb and is saved in .xlsm format to save space and to allow the macros that Dan Mecklenburg at the Ohio DNR developed to be functional.
It is at this point that I’m seeking some help. If anyone has some insight into how to reasonably estimate effective discharge for any of these gauges based on the available data, I would love for you to contact me. Please feel free to use the comment section.
Read Full Post »