xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader 00000nas 2200000Ka 4500
controlfield tag 008 s2005 xx |||| |||||| eng
datafield ind1 7 ind2 024
subfield code a M36-00492-ML-1103
Guswa, Andrew J.
Descarga del Ro Guacimal/Quebrada Cuecha.
Discharge of Ro Guacimal/Quebrada Cuecha.
g 2 de febrero 2005/February 2, 2005.
Books / Reports / Directories
Discharge of Rio Guacimal
Descarga del Rio Guacimal
Scanned by Monteverde Institute.
The State of Water in Monteverde, Costa Rica: A Resource Inventory
PICKER ENGINEERING PROGRAM Northampton, MA 01063 Memo February 2, 2005 From: Andrew J. Guswa, Amy L. Rhodes To: Nat Scrimshaw, Directo r, Monteverde Institute cc: Silvia Newell, Jason Niebler, Ilona Johnson Re: Discharge of Rio G uacimal/Quebrada Cuecha This memo is in response the recent interest in stream discharge of the Quebrada Cuecha (QC) / Rio Guacimal (RG). This document includes the following components: brief overview of streamflow measurement brief chronology of measurements on QC/RG near the cheese factory analysis of data from 2001-2003, recorded on RG about 200 m downstream of the cheese factory analysis of data from 2004, recorded on QC, just upstream of the cheese factory upcoming opportunities Streamflow Determination Determination of discharge in natural st reams is a challenging problem. The most common method of doing so is to create a correlation between stream depth (stage) and the volumetric flux (discharge) of the stream (measured as volume per time, e.g., liters per second). The reason for doing so is that stream depth can be recorded continuously using a pressure transducer and datalogger, while actual measurements discharge are time consuming and require human resources. (Discharge is measured by measuring stream velocity and channel geometry and int egrating to determine the flowrate.) By establishing a correlation between stage and discharge, continuous records streamflow can be inferred from records of stream depth. Such a correlation between stage and discharge will depend on channel morphology, and in a steep, rocky, natural channel (such as QC) there is no a priori expectation of the functional form for this relationship (other than that discharge will increase as stream depth increases). Therefore, many simultaneous measurements of stage and discharge are required to fully express the correlation between these quantities. Of course, if the channel changes shape, a new correlation will need to be established. Additionally, there is no guaranteed way to predict the shape of the correlation below the lowest streamflow measurement or above the largest discharge measurement Direct measurement of discharge when streamflow is low is very challenging. In a rocky channel, it may not be possible to get reliable measurements of velocity and channel geometry during times of low flow.
ANDREW J. GUSWA PICKER ENGINEERING PROGRAM 2/2/2005 page 2 of 5 Because of the challenges associated with working in a natural channel, especially when attempting to determine low flows, it is common practice to use a flow control structure to increase reliability and accuracy. Such structures could be weirs, flumes, sluice gates, or small dams, all of which have the charac teristics of creating a regular geometry for which there exists a theoretical relationship between stream stage and discharge. For example, for flow over a rectangular weir, st ream discharge is related to stream depth by the following relationship: 2 / 3bH C Qd weir where H is the height of the water above the weir, b is the width of the weir, and Cd is a coefficient that needs to be calibrated by a few direct measurements of discharge. History of Measurement on RG/QC January 2001 Prof. Amy Rhodes travel to Montev erde, Costa Rica with Evelyn Kim Â’02, and Sarah Katchpole Â’02 to install a pressure transducer and datalogger on the Rio Guacimal at a location adjacent to the community art center, just downstream of the bridge and cheese factory. The pressure transducer measures water pressure and temperature and converts these meas urements into stream depth; the depths are recorded at regular intervals (10 minutes) by a datalogger, which stores the data until downloaded. The depths recorded by the pressure transducer are corroborated by depths manually read off of a staff gauge that is installed in the stream bed. The reason for this is that the staff gauge is less likely to be disturbed over time and can serve as a reference for the pressure transducer. Summer 2001 Prof. Amy Rhodes returns to Monteverde with Annalee Wells, Eveyln, and Sarah. The team measures discharge on RG200 to establish a correlation between stream depth and discharge. November 2001 The pressure transducer is knock ed over by a large streamflow; therefore, the data beyond this point (until the transducer is repositioned and recalibrated) cannot be used to infer stream discharge. January 2002 Prof. Amy Rhodes returns to Monteverde with Prof. Drew Guswa and Irma Torres Â’04 and Evelyn Kim Â’02. The pressure transducer is repositioned and recalibrated. late Feb 2002 Transducer is knocked over; Stewart Dallas repositioned and recalibrated the instrument within two weeks. late March 2002 The datalogger is unplugged from the art center resulting in a loss of data during the dry season of 2002 (until Amy and Drew return in July). July 2002 Amy, Drew, and their spouses, Erik and Sue, travel to Monteverde. The datalogger is restarted. late Aug. 2002 The transducer is knocked over; Stewart Dallas repositions and recalibrates the instrument within three weeks. June 2003 Amy, Drew, Silvia Newell Â’04 and Aruna Sarma Â’04 travel to Monteverde. Upon arrival, the datalogger is downloaded, and the
ANDREW J. GUSWA PICKER ENGINEERING PROGRAM 2/2/2005 page 3 of 5 recorded stream depths show odd/unusual behavior that started in January 2003. We suspect a failure of the pressure transducer (though the cause of the failure is not clear), and the data are eliminated as unreliable. Again, we have no data from the dry season The transducer is repositioned and recalibrated. January 2004 Amy, Drew, Silvia, and Ilona Johnson Â’06 travel to Monteverde and note that there has been a significant change in the channel morphology at RG200. Specifically, a large debris dam has accumulated just downstream of the pressure transducer resulting in a significant increase in the depth (and width) of the stream at the transducer location. We determine that the stage-discharge relationship determined in prior years is no longer valid. The presence of the collapsed footbridge just upstream of the transducer/datalogger and the threat of imminent changes to channel morphology lead us to seek a new location at which to measure stream discharge. Yet again, we lose the opportunity to determine dry season discharge June 2004 Amy, Drew, Silvia, Ilona, Liz Koeni g Â’05, and Mai Kobayashi Â’06 travel to Monteverde and establish a new gauging site at QC200. The new measurement location is upstream of the cheese factory, and the channel is bounded by rock. Over a period of eight weeks, the students determine a correlation between stream depth and discharge. Fall 2004 In a big event, the precise timing of which is unknown, the staff gauge at QC 200 is moved (knocked sideways). This changes the relationship between discharge and the reading of the staff gauge, though the expectation is that the change will be a simple additive error. January 2004 Silvia Newell continues to measure discharge at QC200 to determine a new relationship between discharge and the staff gauge reading and whether or not the error between the current relationship and that determined over the summer can be corrected by a simple offset. Dry Season Data As indicated from the above chronology, dry season streamflow has been measured reliably only in year 2001. Data from 14 Jan through 24 May 2001 indicate that the flow in the Rio Guacimal at our first gauging site was at or below 57 liters/second for 84% of the time. The rate of 57 liters per second represents the lowest flow ever measured directly, and, thus, provides the lower bound fo r inference of streamflow from stream depth measurements. Records of stream st age indicate that streamflow dropped below this value, but inference of what those streamflows were requires that we extrapolate our correlation between stream stage and discharge. There is no guarantee on the reliability of such an extrapolation Performing this extrapolation indicates that the average flow during the dry season is approx. 10 liters/second (but, again, there is no way to know if such an extrapolation is accurate).
ANDREW J. GUSWA PICKER ENGINEERING PROGRAM 2/2/2005 page 4 of 5 Data from 2004 at QC200 During the summer of 2004, we relocated our stream gauging station to a spot on the Quebrada Cuecha that is upstream of the cheese factory. Over the summer (from June 12 through July 23, 2004) our students, Silvia, Ilona, Liz, and Mai, measured discharge at this location many times per week to uncover the correlation between stream stage (as measured on our staff gauge) and flowrate. The minimum discharge measured was approximately 140 liters per second, and the maximum measured flowrate was approximately 700 liters per second. Our correlation of stage and discharge is applicable within this range. Figure 1 presents a time history of stream discharge at QC200 for the summer of 2004 with these bounds identified by dashed lines. 160 170 180 190 200 210 0 200 400 600 800 1000 1200 Day of Year (2004)Discharge [l/s] Figure 1: Time history of streamflow at QC200 in Monteverde, Costa Rica from June 12 through July 23, 2004. The dashed lines represent the upper and lower bounds of our stage-discharge relationship. Another useful way to look at discharge data is through a flow-duration curve. Such a curve presents discharge versus the fraction of time that the stream had a flowrate larger than that specified. This provides insight into the fraction of time that streamflow might be above or below some threshold level. Figure 2 presents such a curve for the data obtained from June 12 through July 23, 2004. As can be seen in the figure, streamflow is rarely below 150 l/s or above 300 l/s during this part of the wet season.
ANDREW J. GUSWA PICKER ENGINEERING PROGRAM 2/2/2005 page 5 of 5 0 0.2 0.4 0.6 0.8 1 150 200 250 300 350 400 450 500 550 600 650 700 Fraction of Time ExceededFlowrate [l/s] Figure 2: Instantaneous flow-duration curve for QC200 in Monteverde, Costa Rica for streamflows from June 12 through July 23, 2004. Opportunities Reflecting on the limitations of our collected data and the needs of the community, there is clearly an opportunity this winter a nd spring to improve our understanding of dryseason streamflow of the Rio Guacimal / Quebrada Cuecha. We have left equipment to measure discharge at the Monteverde Institute, and Silvia Newell is well-versed in making such measurements. Regular and cons istent measurements of streamflow will strengthen our confidence in the current correlation between stream stage and discharge expand the correlation to flows not currently bounded by our measurements (especially on the low end) provide a baseline of streamflow data even in the event that the continuous record of stream depth is disrupted (the pressure transducer fails or the water level drops below the transducer) To further extend the utility of these data, we recommend that measurements of discharge also be made at RG100. This will provide insight into how much (what fraction) of the flow for the larger watershed (as measured at RG100) is contributed by the Quebrada Cuecha.