Marcell Experimental Forest daily streamflow data

Metadata:

Data_Quality_Information:

Attribute_Accuracy:

Attribute_Accuracy_Report:

Descriptions of the accuracy of measurement of attribute values can be found in the Methodology section for each attribute measured. Streamflow is calculated from stage-discharge relationships for H-type flumes or v-notch weirs. Data resolution is considered to be 5% of total flow.

Additional Information:

The flumes at S1, S4N, S4S, and S5, where there were cutoff walls, were prone to occasional leaking under the walls. The flume approaches, which were supported atop small concrete pads in channels downstream of the walls, were also prone frost-heaving and settling. The flumes may also have frozen when propane heaters malfunctioned, causing icing in the approach and H-type facing (\Supplements\Photos\S4_xxxx_cutting_flow_path.jpg), or freezing of floats in flume stilling wells. The flume structures were closely monitored during weekly or more frequent site visits to allow for the adjustment of flume approach elevations to maintain levelness, remove ice, or free floats. For these reasons and because flumes were less suitable than v-notch weirs for the long periods of low flow that occur at the MEF research watersheds, the flumes were eventually replaced with weirs, usually after a period of repeated breaches. During the early 1960s, the S2 weir was also prone to ice formation in the v-notch until an insulated wooden shelter with a heater was placed over the weir and v-notch before first freeze and removed after spring snowmelt. Now, all weirs are similarly covered and heated. If streamflow stopped and water levels dropped below a v-notch, a heater was turned off to conserve fuel until flow resumed. With flow resumption, a propane lamp was relit to prevent ice formation in a v-notch. The flumes were also enclosed and heated after the first several years of operation during the 1960s. Stages measurements were also affected by occasionally damning of flumes by beaver (Castor Canadensis), which was usually resolved by trapping and removal. Stream stages were adjusted for all periods in which it was known that stage measurements were affected.

During brief periods of streamflow recession and no rainfall or snow melt, gaps in stream stage records could be estimated by forecasting or backcasting the recession through the gap. A note on a stripchart (April 13-20, 1977) provides some additionally insight on how missing data during periods of stormflow were estimated for the S4S stream gage. “This is an estimate using the precip record, S4N and S4 recording well chart,” by Art Elling on April 22, 1977. In practice, that meant that stage at the S4N weir was used to determine the timing and magnitude of changes in stream stage that would have occurred at S4S. The precipitation record, as recorded on a hygrothermograph at the North Meteorological Station, was used to validate the occurrence of rainfall events, which may have led to increased stream stage. Peatland water table elevation, as recorded at the recording peatland well in the S4 bog, was used as an additional validation approach by comparing the stream stage at the S4N and the reconstruction for S4S to the timing and magnitude of changes in water table elevation in the bog upstream of the weirs. Correlations of streamflows from two stream gages before and after missing/incorrect periods were used to estimate streamflow during data gaps. Generally, stream stage increased proportionally to increases in peatland water table elevation. This type of approach was used to correct the magnitude of stream stage for all stream gages.

The highest flow of 608.8 liters per second (L/s) at the S5 weir occurred when water overtopped the weir. The recorder indicated depth over the wall, which allowed flow estimation by calculating the area of the cross section over the wall and multiplying by velocity measurement from a Price Pygmy Current Meter.

Streamflow at S6 was measured with an H-type flume with concrete cutoff walls from March 1963 to September 1974. The flume structure was replaced with a V-notch weir and streamflow measurements resumed during 1976. The weir blade was set 7.6 centimeters (cm) too high causing water to occasionally back up into the bog until the weir blade height was corrected on July 1, 1977. The mistake went unnoticed until June 1977, when the v-notch was repositioned to the elevation of the original stream contour. Streamflow was measured accurately while the weir blade was set too high and that mistake likely has minimal effects on the interpretation of S6 streamflow data because there was little streamflow outside of snow melt that year, and the maximum stage height of 0.33 inches (10 cm) was not enough to back standing water into the bog during a period of high evapotranspiration. In contrast, the change from a flume to a weir may have had an important bog-scale affect on water partitioning to streamflow and evapotranspiration. The change from a rectangular geometry of a flume to a v-shaped geometry caused higher water levels that backed into the peatland during high flows. The feedbacks between water level, streamflow, evaporation, and transpiration are complex and difficult to disentangle because evaporation and transpiration in the bog are sensitive to water level (Nichols and Brown 1980; Brooks et al. 2011). The stream stage data were compiled and have been used to estimate streamflow. Though the data were shared in the past, we have not released those data in this data publication because the effects on streamflow are not consistent with the primary reasons (small catchment and paired-catchment studies) for the streamflow record or for hydrological-process comparisons of the post-1976 to the pre-1974 data.

All years of stage data for all stream gages were recompiled and streamflow was recalculated during 2016. Changes to streamflow values from the previous compilations/calculations were common, though usually trivial in magnitude unless mistakes were identified in the past digitization/recording of values for particular days and stream gages. The slight variations reflect improved computational power of desktop computers, refined coding, and removal of errors upon identification. The values reported in this data archive should be used hereafter and previous versions should not be considered.

This data publication containing daily streamflow record does not currently indicate the days on which stream stage was estimated. Stripcharts are available to do so if resources become available in the future to allow such documentation to occur.

Nichols, Dale S.; Brown, James M. 1980. Evaporation from a Sphagnum moss surface. Journal of Hydrology 48(3-4): 289-302. https://doi.org/10.1016/0022-1694(80)90121-3

Brooks, Kenneth N.; Verma, Shashi B.; Kim, Joon; Verry, Elon Sandy. 2011. Scaling up evapotranspiration estimates from process studies to watersheds. Pages 177-192. In: Kolka, Randall K. (Ed.); Sebestyen, Stephen (Ed.); Verry, Elron S. (Ed.); Brooks, Kenneth N. (Ed.). Peatland biogeochemistry and watershed hydrology at the Marcell Experimental Forest. CRC Press, Boca Raton, FL. https://doi.org/10.1201/b10708-7

Logical_Consistency_Report:

not applicable

Completeness_Report:

The headwater streams draining the MEF research watersheds may have prolonged periods of no flow, especially during winters when most water that would otherwise be available to streamflow is frozen. Once flow stops in winter, stripchart recorders are not operated until a weather forecast indicates a period of above freezing temperatures that may induce flow to resume. Therefore, there are sometimes no stripcharts for weeks of no flow. The last and first charts of the season were notated on stripcharts.

Some stage measurements are not available due to equipment malfunctions. Recorder malfunctions are uncommon, but when they happen, stream stage was estimated from precipitation records, bogwell records and streamflow from other watersheds. Recorder malfunctions included: incomplete records when clock-drives stopped or time-distorted records when clocks sped up or slowed down; indefinite marking due to ink splotches; incomplete records due to unreadable or incomplete ink traces on paper stripcharts; shifted pen markings of stage height that occurred when the pen or pen-drive was not working properly; and flat-lined or missing traces when float tapes jumped from pulleys.

There were some prolonged periods during which stream stages were not recorded at stream gages or otherwise not accurately recorded that are noted here:

S1: Streamflow was measured occasionally at a temporary wooden flume during 1960 (\Supplements\Photos\S1_1960_temp_flume.jpg) and an H-type flume with wood cutoff walls (\Supplements\Photos\S1_1961_flume_wood_cutoff_wall.jpg) was installed later that year. The whereabouts of those data currently are not known. After the installation of a flume, ice heaving of the approach and cutoff walls was an ongoing problem. The cutoff walls heaved during the winter of 1970 and streamflow measurements were curtailed until May 1970 when the walls were again driven into a cemented “ortstein” sand layer. The wood cutoff walls were replaced by manual driving of interlocking steel sheetpiling during October 1974 (\Supplements\Photos\S1_1974_sheet_pile_driving.jpg). Ice damage continued to be a problem for upkeep of the flume until streamflow measurements were discontinued and the flume was removed during December 1980. The metal cutoff walls remained in place until the aboveground portion was removed during the winter of 2015.

S2: A 120-degree v-notch weir was constructed and stripchart recording was implemented during October 1960. (See: \Supplements\Construction_plans_S2_1960.pdf for information about the construction of the weir, and \Supplements\Photos\S2_1960_weir_construction.jpg, \Supplements\Photos\S2_1960_weir_pond_filling.jpg for images of the weir). Under low flow during winter, the v-notch was prone to severe freezing (\Supplements\Photos\S2_1960_weir_frozen.jpg), a problem that was later alleviated by placing a shelter with a propane heater over the v-notch and concrete walls (\Supplements\Photos\S2_2014_weir.jpg). On July 6, 1983, an earthen dam was placed upstream of the weir pond to prepare for replacement of the original stream gage. The original weir blade was removed from the concrete walls and placed in the earthen dam, where occasional measurements of stage were recorded until July 7 when flow stopped. These stage measurements were used to estimate stage height until the current V-notch weir (\Supplements\Photos\S2_1983_weir.jpg, \Supplements\Photos\S2_2015_weir.jpg) and concrete cutoff walls were completed on 18 July 1983.

S3: Monthly and annual estimates of S3 streamflow, as described in Sebestyen et al. 2011 are not yet available, but will be added to this publication in the future. The S3 peatland is part of a larger peatland complex. There is another peatland immediately downstream of the narrow outlet area, where there was not a defined, stable channel prior to 1967. To create a channel in which to measure surface water flow, vegetation was cleared along a 30-meter-long path downgradient from the S3 outlet and this 3- to 5-meter-wide strip was ditched using dynamite during the summer of 1967 (\Supplements\Photos\S3_1967_outlet_ditching_1.jpg, \Supplements\Photos\S3_1967_outlet_ditching_2.jpg, and \Supplements\Photos\S3_xxxx_veg_clearing.jpg). Plywood cutoff walls were erected to form a stable control area within the ditch (\Supplements\Photos\S3_xxxx_plywood_cutoff_walls.jpg). Water velocity and level were measured occasionally to develop a predictive equation for streamflow using the regression between water stage and discharge. Water level in the control section at the walls was recorded on stripcharts during the ice-free period from 1967 to 1975. Since this calibration period, the stream has not been gaged because beaver repeatedly built dams downstream of S3. The dams elevated water levels in the S3 drainage ditch, which changed the stage-discharge relationship during those periods. A regression between water level at the fen well in the S3 peatland and streamflow has been used to calculate monthly and annual water yields from S3 from 1963 to the present. These estimates of streamflow have not yet been released. We anticipate doing so, contingent upon future refinement of the estimation approach, validation, and complete documentation of the process.

S4: The S4 watershed sits atop the continental divide of the Mississippi and Hudson Bay drainages. Surface water flows from two outlets at 428 meters (m) above mean sea level and streamflow is measured at both outlets. About 70% of the stream water flows from the north outlet (stream gage S4N) to the Hudson Bay drainage with the rest flowing through the south outlet (stream gage S4S) to the Mississippi River (Verry 1972). Streamflow from the two stream gages is summed to report streamflow for the entire S4 watershed.

Streamflow at S4N was measured with an H-type flume with cutoff walls made of metal sheet piling from December 1961 to 1984. The flume began leaking during May 1984, with most flow leaking underneath the cutoff walls. Stream stage was not recorded from May 8 through November 9, 1984. During November 1984, the complete flume/wall structure was removed and a 120-degree v-notch weir with concrete cutoff walls was constructed. For the period of missing data during 1984, streamflow appears to have been estimated as ~60% of streamflow at the S5 watershed. The period from March 1984 to February 1985 needs to be recalculated and that will be done once the stripcharts are re-digitized. The original stripchart scans are no longer available. Future updates to daily streamflow values should not be considerably different from values reported in this data archive, and are not expected to greatly affect annual streamflow totals or interpretation of annual or longer summations of streamflow.

The stream gage at S4S had flume walls and a short approach section that was identical the flumes at the S1, S4N, and S5 stream gages. Below the approach section, a V-notch weir plate was attached to the front of a lower-level flume box rather than having an H-type facing. The cutoff walls were constructed of metal sheet piling. The S4S gage was reconfigured as an H-type flume during 1965. During 1980, the entire S4S flume/wall structure was replaced with a V-notch weir with concrete cutoff walls. On October 14, 1980, the flume was removed from the wall during a low flow period, the opening was dammed with clay, and the replacement weir was constructed between October 14 and 30. The weir was placed at about the location of the H-type facing. There was no flow through the stream gage until November 1 when the old wall was removed, water filled the weir pond, and flow through the 120-degree opening of the v-notch weir started. From November 1-3, stream stage was estimated from occasional observations until the stripchart recorder was installed on November 3.

Users of the S4S data between 2002 and 2010 may have been provided incorrect streamflow values. During 2002, stripchart digitization was started and Visual Basic code was used aggregate stage data and calculate streamflow. The incorrect streamflow values stemmed from the use of the H-flume rating during the period from April 11, 1962 to April 16, 1965 when a V-notch weir rating should have been used. During 2010, the calculation error was discovered. The data have been recalculated using the appropriate v-notch weir rating for that period. We are not aware of publications reporting the streamflow at S4S during the 2002-2010 period when that error was not recognized.

S5: Streamflow at S5 was measured with an H-type flume with cutoff walls made of metal sheet piling from March 1962 to September 1982. On September 9, 1982, the flume and approach were removed from the sheet piling during a time of no flow. A 120-degree v-notch weir with concrete cutoff walls was constructed downstream of the walls. There was no streamflow during the period of construction. Stage height measurements resumed on September 21 when the walls were removed, releasing water that was pooled behind the walls and allowing streamflow to resume.

Vandalism on October 11, 2011 resulted in a 15-day period of no recorded stream stage at S5. Stream stage recording was re-established on October 26 after damage to the gage housing and culvert were repaired. There was no streamflow on October 11 and flow resumed on October 18. To estimate stream stage, height of water in the v-notch was measured on four dates during site visits. Streamflow was low (maximum of 0.5 L/s) throughout the period when stage was not recorded.

S6: Streamflow at S6 was measured with an H-type flume with concrete cutoff walls from March 1963 to September 1974. A crack in the cement wall was found and repaired during autumn 1973. Winter frost heaved the concrete structure during February 1974 creating a crack that was deemed unrepairable. The flume structure was replaced with a V-notch weir and streamflow measurements resumed on March 1, 1976. The weir blade was set 7.6 cm too high causing water to occasionally back up into the bog until the weir blade height was corrected on July 1, 1977. Water levels were low during the drought of 1976 and the mistake went unnoticed until June 1977 when water levels rose after the drought. Even after correct placement to the depth of the original stream contour, the change from a rectangular geometry of a flume to a v-shaped geometry caused higher water levels that backed into the peatland during high flows. The stream stage data were compiled and have been used to calculate streamflow. Though the data were shared in the past, we have not released those data in this data archive because the change in stream gaging structure makes comparison of the post-1976 to the pre-1974 data impossible because evaporation and transpiration in the bog are sensitive to water level (Nichols and Brown 1980; Brooks et al, 2011) and would feedback on streamflow amounts. We recommend that the 1963-1974 data not be considered in the context of small catchment or paired-watershed studies.


Verry, Elon Sandy. 1972. Effect of an aspen clearcutting on water yield and quality in northern Minnesota. In: Proceedings of a symposium on Watersheds in Transition held in Fort Collins, Colorado, June 19-22, 1972, edited by S. C. Csallany, et al., pp. 276-284, American Water Resources Association, Urbana, IL.

Lineage:

Methodology:

Methodology_Type: Field

Methodology_Description:

Streamflow:

Streamflow has been measured at as many as seven stream gages. Streamflow currently is measured at five 120-degree v-notch weirs that replaced earlier H-type flumes or weirs. The v-notch and flume bottoms were set to the elevations of the stream channels that drained bogs. Pools were excavated behind the stream gages and channels were contoured downstream to create a hydraulic drop. Stream gages at the outlets of the S2 and S4 watersheds are downstream from the peatlands. By contrast, stream gages and weir pools at the outlets of the S5 and S6 watersheds are adjacent to the bogs and pooled water backs into the bogs as water levels rise.

Stream Gages:

The watersheds have single channel outlets where streamflow is monitored, except the S4 watershed, which has 2 outlets (S4N and S4S). The S4 watershed sits atop the continental divide of the Mississippi and Hudson Bay drainages. Surface water flows from two outlets at 428 m above mean sea level and streamflow is measured at both outlets. Streamflow from the two stream gages is summed to report streamflow for the entire S4 watershed.

The S2 stream gage was always a v-notch weir with concrete cutoff walls and the S1 stream was always a flume. Before being replaced with v-notch weirs and concrete cutoff walls, the stream gages at S4N, S4S, and S5 were configured as H-type flumes with metal cutoff walls from the 1960s to 1980s. The S6 flume was replaced with a v-notch weir during the 1970s.

Each stream gage now has a 120° v-notch weir blade, constructed of ¼ inch steel. The blades are fastened to concrete cutoff walls. The walls are 10 inches thick and 7 feet (ft) high. Most of the wall is buried in the soil and sits on an 18 inch x 36 inch footing. The walls were constructed perpendicular to the natural channels and are long enough to completely block the flow and force it through the weir. Total length of the wall varies from 18 ft to 30 ft, depending on the shape of the flood prone area. The weir was placed so the bottom of the V was at the level of the natural stream bottom. A pool approximately 3 ft deep and 15 ft in diameter was shaped behind the wall to provide an area for the flow to be dispersed to prevent a current through the pool and over the weir. The footing and wall were attached with ½ inch rebar, 1 ft on center and a 1½ inch x 3½ inch keyway to prevent seepage between the wall and footing. The wall was backfilled with clay soil to assure no seepage around or under the structure. The stilling well at S2 was poured concrete with a wooden gage house built atop. The stilling wells at other weirs were made from galvanized steel culvert, with a wooden shelter placed atop to house a stripchart recorder. A 1¼ inch pipe connects a stilling well to a pond center.

Prior to replacement with weirs, the S1, S4N, S4S, S5, and S6 stream gages were 1.5 ft (45 cm) type-H flumes with cutoff walls (\Supplements\Flume_drawings_1960.pdf). The walls were constructed perpendicular to the natural channels. The bottom of the inlet to the flume approach (at the wall) was placed at the level of the natural stream bottom. The flume approach was ~3 m long and supported atop small concrete pads or cinder blocks in channels downstream of the walls. A pool was shaped behind the wall to provide an area for the flow to be dispersed to prevent a current through flume. A stilling well was attached to H-type facing of the flume and a galvanized steel metal enclosure was placed atop each stilling well.

Elevations of the "0" mark on the point gauge, the weir notch, and the reference point for the stripchart recording in the gage house were checked yearly from a known benchmark.

After the first several years of operation and the severe effects of ice-damming became apparent, propane lamps were placed inside the flume approaches to eliminate freezing. Now, all weirs are seasonally covered and heated. If streamflow stops and water levels drop below the v-notch, a heater is turned off to conserve fuel until flow resumes. At that time, a propane lamp is relit to prevent ice formation in a v-notch. Weir covers are removed in the spring when streamflow is high enough to prevent freezing in the notch.

During ice free periods, screened boxes were placed over the v-notch weir blade of the flumes to keep floating debris from clogging the v-notch.


Data Collection:

Though many individuals have assisted with field measurements, stream gages and records have been maintained by remarkably few hydrological technicians over the entire duration of the streamflow records. Clarence Hawkinson worked from 1960 until 1974, Art Elling from 1969 to 2005, Richard “Deacon” Kyllander from 1981 to present, Carrie Dorrance from 2005 to 2017, and Tyler Roman from 2017 to present. The technicians were responsible for field measurements, infrastructure and instrument maintenance, data recording, and data curation.

Stream stage has been recorded with a Stevens Water Monitoring Systems, Inc (Portland, OR) Type A-35 stripchart recorder, with a precision of 0.3 cm, at the S2 weir. Stream stages at all other stream gages has been recorded with Belfort (Baltimore, MD) FW-1 stripchart recorders, with a precision of 0.6 cm. The recorders are mounted above the float stilling wells and housed inside plywood shelters. The float, tape, and pulley system is located inside a stilling well at each stream gage and the recorded stage corresponds to water height in the weir notch (or flume). Charts on the FW-1 recorders are changed weekly and the A-35 charts are checked weekly, but retrieved about every 2 months. Pulleys on stripchart recorders are driven by a float, tape, and counterweight systems. The A-35 has a 1.5 ft (45.7 cm) circumference pulley, and the FW-1 recorders have 1 ft (30.5 cm) circumferences pulleys. The FW-1 recorder advances the stripchart at 4.6 centimeters/day (cm/d) with 6 hour (h) increments between major divisions on the chart vertical axis. The Type A recorders advance a stripchart at 12.2 cm/d with 6 h increments between major divisions on the chart vertical axis. Propane heaters were placed inside the stilling wells to maintain open water when air temperatures are below freezing.

Stream stage height was checked with a weekly reading of an aluminum point gauge (Brakensiek et al. 1979) on the outside of each gage shelter, when a pool was ice-free. The height of recorder pens were adjusted as needed to match stage recorded on the stripchart to stage measured with the point gage. Adjustments to pens and stage heights were noted on stripcharts. The date/times, measured stage from the point-gage, and notes were written on a stripchart weekly, or when a stripchart was placed on and taken off a stripchart recorder. Recorders are cleaned and lubricated on a yearly basis, or more frequently if needed.

In the Grand Rapids Forestry Sciences Lab, stripcharts are marked with dates across the top of the chart, a V- or check mark at the intersection of pen trace and midnight time lines. Corrections to time and stage were directly annotated on stripcharts, as per the Standard Operating Procedure document (\Supplements\SOPs_MEF_1961.pdf). Chart readings were adjusted to match the point gauge (Brakensiek et al. 1979) reading. Recorder malfunctions are uncommon, but when they happen flow values are estimated from precipitation records, bogwell records and stream runoff from other watersheds.

The corrected values on stripcharts were then digitized from scanned stripcharts. Digitization instructions are included as a supplement (\Supplements\Chart_digitizing_instructions_2008.pdf). Stage data were digitized as sub-daily breakpoint data (Johnson and Dils, 1956) in a spreadsheet file (.csv format) for each day of the year when streamflow occurred. At a minimum, stage was recorded at 0:00 and 24:00. During stormflow, the time and stage at every inflection point on the hydrograph was digitized. The sub-daily files for a water year (March 1 to Feb 28) were compiled into a spreadsheet (.xlsv format) using macros (VBA, Visual Basic) in Microsoft Excel. The VBA code includes a formula to correct for the curved-time lines on FW-1 stripcharts. Streamflow is calculated for each stage measurement using stage-discharge relationships in a spreadsheet. The areas in polygons that corresponded to each breakpoint data interval were calculated to determine the streamflow per time interval. Sub-daily estimates of streamflow were summed for each day and daily streamflow is reported in units of cubic feet per second (CFS), centimeters per day (cm/d), and liters per second (L/s).

Streamflow characteristics:

Streamflow from the S1, S2, S4, S5, and S6 watersheds is relatively low, with periods of no flow during all years on record. For example, there was no flow for 38% of the S2 streamflow record (Amatya et al. 2016). Peak flows are generally less than 56.6 liters per second (LPS), but a large storm in September 1988 produced the highest flow of 608.8 LPS at the S5 weir. Flows typically start in late March or early April when snow melt occurs. Streamflow is typically elevated for a prolonged period after snow melt, yet nine of ten of the largest streamflow events have occurred during rainfall runoff events, not snowmelt or rain-on-snow events. Streamflow may occur throughout the summer, but oftentimes stops during dry spells. Usually, streamflow will start again with significant rains and then continue until early or mid-winter when they stop again until the spring snowmelt event.

Streamflow dynamics at S3 are quite different due to the presence of a fen in this watershed versus bogs in the S1, S2, S4, S5, and S6 watersheds. The S3 fen and stream are fed by groundwater discharge. While peak flows are the same between the bog and fen watersheds, low flows are much higher in the S3 stream larger and periods of no flow are less common there than in the bog watersheds (Verry et al. 2011).

Data Format:

Daily streamflow data from the S2 weir are reported from 1961 through 2017. Daily data from the S4S, S4N and S5 weirs are reported from 1962 through 2017. Daily data from the S6 weir are reported 1976 through 2017.

Methodology_Citation:

Citation_Information:

Originator: Brakensiek, Donald L.

Originator: Osborn, Herbert B.

Originator: Rawls, Walter J.

Publication_Date: 1979

Title:

Field manual for research in agricultural hydrology

Geospatial_Data_Presentation_Form: document

Series_Information:

Series_Name: Agriculture handbook

Publication_Information:

Publication_Place: Washington, DC

Publisher: U.S. Department of Agriculture

Other_Citation_Details:

550 pp

Methodology_Citation:

Citation_Information:

Originator: Johnson, Edward A.

Originator: Dils, Robert E.

Publication_Date: 1956

Title:

Outline for compiling precipitation, runoff, and ground water data from small watersheds

Geospatial_Data_Presentation_Form: document

Series_Information:

Series_Name: Station Paper

Publication_Information:

Publication_Place: Asheville, NC

Publisher: Southeastern Forest Experiment Station

Other_Citation_Details:

40 pp

Process_Step:

Process_Description:

Streamflow

Data Collection and Conversion:

In the Grand Rapids Forestry Sciences Lab, corrections to time and stage (to weekly point gage measurements; Brakensiek et al. 1979) were directly annotated on stripcharts. Starting during 2002, stripcharts were scanned using a SummaSketch III Professional Digitizing Pad and corrected values on stripcharts were digitized using Sigma Scan software (Copyright © Systat Software, Inc ). Sub-daily stage data were saved as text (.prn or .csv) files for each day. The sub-daily files for a water year (March 1 to Feb 28 or 29) were compiled into a spreadsheet (.xlsv format) using macros (VBA, Visual Basic) in Microsoft Excel and streamflow was calculated for each stage measurement using stage-discharge relationships. The areas in polygons that corresponded to each breakpoint data interval were calculated to determine the streamflow per time interval. Sub-daily estimates of streamflow were summed for each day and daily specific discharge is reported in units of cubic feet per second (CFS), centimeters per day (cm/d), and liters per second (L/s).

Here are the equations used for the conversions:
CFS = (L/s) * 0.035315
L/s = (CFS) * 28.31685
cm/d = (L/s) / (0.864 * watershed area in hectares)

Stripchart and Data Archiving:

Stripcharts are stored in fireproof vaults in a decidated archive room at the Forestry Sciences Lab in Grand Rapids, MN. All computers files are stored with multiple redundancy on desktop computers and backup drives at the Forestry Sciences Laboratory and elsewhere. VBA code was written by Sebestyen and is stored with multiple redundancy on desktop computers and backup drives at the Forestry Sciences Laboratory.

Process Date:

Stream stages were adjusted or estimated for all periods in which it was known that stage measurements were affected by infrastructure maintenance, instrument failure, datum corrections, flow blockages, or ice damming. Stripcharts have generally been marked within a week of removal from a stripchart recorder and checked within a year of marking. The process of digitizing stripcharts began during 2002, and continues to present. In recent years, stripcharts have typically been digitized within a year of collection.

Process_Date: Unknown

Entity_and_Attribute_Information:

Overview_Description:

Entity_and_Attribute_Overview:

Daily streamflow = data containing average daily streamflow for the specified stream gage in the Marcell Experimental Forest. Measurements began in the early 1960's for all of these weirs.

Date = Date of observation. The date format is yyyy-mm-dd.

Watershed = Text identifier for the location of the observation (S1, S2, S4N, S4S, S5, S6).

Flow (CFS) = daily streamflow in cubic feet per second (CFS)

Flow (cm/d) = daily streamflow in centimeters per data (cm/d)

Flow (L/s) = daily streamflow in liters per second (L/s)

Entity_and_Attribute_Detail_Citation:

none provided

Overview_Description:

Entity_and_Attribute_Overview:

The following supplemental files are also available for download:

\Supplements\:

Chart_digitizing_instructions_2008.pdf: Adobe Acrobat PDF file containing digitization instructions (converting strip chart data to electronic data) used for Marcell Experimental Forest strip charts, dated 8/11/2008.

Construction_plans_S2_1960.pdf: Adobe Acrobat PDF file containing the drawings and notes for the 1960 weir.

Flume_drawings_1960.pdf: Adobe Acrobat PDF file containing the drawings and notes for the 1960 flumes.

MEF_map.jpg: JPEG image file containing a map of the Marcell Experimental Forest which includes the location of the watersheds.

SOPs_MEF_1961.pdf: Adobe Acrobat PDF file containing "Standard Operating Procedures for compiling climate, runoff, and ground water data" by Roger Bay, Research Hydrologist, in 1961.


\Supplements\Photos\:

S1_1960_temp_flume.jpg: JPEG image file containing a photo of the S1 temporary flume, taken on 9/9/1960 by Roger Bay (USDA Forest Service). View is towards west or northwest.

S1_1961_flume_wood_cutoff_wall.jpg: JPEG image file containing a photo taken on 9/13/1961 of an H-type flume with wood cutoff walls that was installed after the S1 temporary flume (later in 1960).

S1_1974_sheet_pile_driving.jpg: JPEG image file containing a photo from 1974 of the sheet piling being driven. Sandy Verry is using the sledge hammer. Art Elling is holding the batter plate. View is towards the northeast. Forest Service photography, though person is not known.


S2_1960_weir_construction.jpg: JPEG image file containing a photo of the construction of the S2 weir pond, taken on 10/3/1960. Roger Bay was the photographer, view looking slightly north of west. Person in picture is not known.

S2_1960_weir_pond_filling.jpg: JPEG image file containing a photo of the S2 weir pond filling up. Roger Bay was the photographer, view looking slightly north of west. Person in picture is not known.

S2_1983_weir.jpg: JPEG image file containing a photo of the S2 weir, taken during July 1983. The weir pool is filling after reconstruction of the weir, stilling well, and gage house. Unknown photographer, view looking west.


S2_1960_weir_frozen.jpg: JPEG image file containing a photo of the S2 weir with severe ice buildup, taken on 12/14/1960. Roger Bay, Research Hydrologyist, was the photographer, view looking east.

S2_2014_weir.jpg: JPEG image file containing a photo of the wooden shelter with tanks that provide propane to heat lamps that are inside of the shelter, taken on 4/9/2014. Photographer was Stephen Sebestyen, Research Hydrologist, view looking east. Similar shelters are placed over weirs on streams draining the other research watersheds.

S2_xxxx_weir_1.jpg: JPEG image file containing an image of the v-notch weir in the S2 watershed.

S2_xxxx_weir_2.jpg: JPEG image file containing an image of Art Elling, hydrologic technician, measuring streamflow in the S2 watershed.

S2_2015_weir.jpg: JPEG image file containing a photo of the S2 weir taken on 5/12/2015. Note the wire-mesh box that is placed over the v-notch to prevent floating litter and detritus from clogging the v. Photographer was Stephen Sebestyen, Research Hydrologist, view looking east.


S3_1967_outlet_ditching_1.jpg: JPEG image file containing a photo showing ditching using dynamite along a 30-meter-long path downgradient from the S3 outlet, taken during the summer of 1967. The ditching created a channel in which to measure stream stage. Unknown photographer.

S3_1967_outlet_ditching_2.jpg: JPEG image file containing a photo showing ditching using dynamite along a 30-meter-long path downgradient from the S3 outlet, taken during the summer of 1967. The ditching created a channel in which to measure stream stage. Unknown photographer.

S3_xxxx_plywood_cutoff_walls.jpg: JPEG image file containing a photo of plywood cutoff walls that were erected to form a stable control area within the outlet ditch of the S3 watershed. Date and photographer are unknown. View is looking west.

S3_xxxx_veg_clearing.jpg: JPEG image file containing a photo showing the vegetation being cleared along a 30-meter-long path downgradient from the S3 outlet and this 3- to 5-meter-wide strip was ditched using dynamite during the summer of 1967. Unknown photographer.


S4_xxxx_cutting_flow_path.jpg: JPEG image file containing a photo of the S4S or S4N flume showing severe icing in the flume approach. The chainsaw was used to cut a flow path through ice to restore operation of the flume. Date, photographer, and person pictured are not known.

S4_xxxx_flume.jpg: JPEG image file containing a photo of Hydrological technical Clarence Hawkinson standing next to the stilling well on the S4S weir. The photo was taken sometime from 1961-1965 when the stream gage at S4S was configured as a v-notch weir. Roger Bay is the photographer, view looking north.


\Supplements\Instrumentation\:

S#Instrumentation.pdf: Adobe Acrobat PDF/a files containing a description of instrumentation for the specified S# watershed (where # = 1, 2, 3, 4, 5, or 6) including measurement, type of instrument, record period, and some additional notes.

S#InstrumentationMap.jpg: JPEG image files containing a map of the specified S# watershed (where # = 1, 2, 3, 4, 5, or 6) providing locations of the different instrumentation available (e.g., air temperature, soil temperature, peatland wells, groundwater wells, runoff collectors, rain gages, soil water tubes, weirs, snowcourse location, etc.).

Entity_and_Attribute_Detail_Citation:

not applicable

Metadata_Reference_Information:

Metadata_Date: 20180430

Metadata_Contact:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Dr. Randy Kolka or Stephen Sebestyen

Contact_Organization: USDA Forest Service, North Central Research Station

Contact_Address:

Address_Type: mailing and physical

Address: Forestry Sciences Lab

Address: 1831 Highway 169 East

City: Grand Rapids

State_or_Province: MN

Postal_Code: 55744

Country: USA

Contact_Voice_Telephone: Randy: 218-326-7100, Stephen: 218-326-7108

Contact_Electronic_Mail_Address: Randy: rkolka@fs.fed.us, Stephen: ssebestyen@fs.fed.us

Metadata_Standard_Name: FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata

Metadata_Standard_Version: FGDC-STD-001.1-1999