1. Data Set Overview 2. Investigator(s) 3. Theory of Measurements 4. Equipment 5. Site Description 6. Data Acquisition Methods 7. Observations 8. Data Description 9. Data Manipulations 10 Errors and Limitations 11. Software 12. References 13. Glossary of Terms and Acronyms 14. Document Information 1. Data Set Overview 1.1 Data Set Identification [This section is aimed toward the person searching for a data set. Descriptions should be succinct and clear, and acronyms should be spelled out.] Boreal Ecosystem Research and Monitoring Sites (BERMS) Tower Flux Meteorological Data From the Southern Study Area Old Black Spruce Site 1.2 Study Overview [A short text describing the study/experiment, and its objective] The BERMS (Boreal Ecosystem Research and Monitoring Sites) project was designed as a 5 year partial follow-up to the BOREAS (Boreal Ecosystem-Atmosphere Study) experiment. The data set documented here, includes the near-surface meteorological measurements in support of carbon, water and energy flux measurements at the Old Black Spruce site in Saskatchewan. The climate monitoring program serves some of the following purposes: 1) to provide supporting measurements for flux monitoring, 2) to contribute to the development and validation of weather and climate models and 3) to provide information for interannual climate variability studies in the boreal forest. BERMS has been a participant in Fluxnet Canada since 2002. 1.3 Data Set Introduction [The nature of the data, including a summary of the key parameters/variables studied, and the primary instruments used. A full description will be given in section 7.] The main meteorological data set is in text format. Variables have been delimited by commas, to make it easy to import into most software programs. Files have been grouped by site, month and then by data type. For example, at Old Black Spruce in one month, there are three files: the first contains ANCILLARY data (extra data not normally required by users), the second contains the MAIN subset of measured variables and the third is includes SUMMARIZED and gapfilled data. These files have two header lines, the first for variable names in APL (Adjective_Parameter_Location) format, and the second for units. Here is a summary of the measured parameters (most are measured at 30min intervals): -Air temperature: measured as deg C at various heights in relation to the structure of the forest canopy. -Humidity: measured as relative humidity at various heights in relation to the structure of the forest canopy. -Precipitation: snow depth, precipitation accumulation, rain events measured in mm. -Wind direction: measured in compass degrees above the forest canopy. -Wind speed: measured in meters per s above and within the forest canopy. -Atmospheric pressure: surface pressure measured in millibars (or Pascals). -Soil temperature: measured in deg C at 1 or more locations and at various depths in relation to ground level. -Soil moisture: volumetric water content measured at 1 or 2 locations and at various depths in relation to ground level. -Radiation: measured in watts per m sq. Includes: net radiation, downwelling shortwave, upwelling shortwave, downwelling longwave, upwelling longwave, diffuse shortwave radiation, global solar radiation and photosynthetically active radiation. -Water table height: measured in mm. 1.4 Related Data Sets [Note any similar or related data collected by the investigator, other investigators, or other data centres. Something like five or six related data sets is a good number to provide.] 1) BOREAS SRC mesonet climate data available from 1993 to late 1996. 2) CO2, CH4, CO concentration data measured by Air Quality Division of the Climate and Atmospheric Research Directorate of the Meteorological Service of Canada of Enviornment Canada 3) Extra treebole temperature data (2-3 trees per site) 4) Kari Bisbee’s 3 surface met stations at OBS. 5) Ken van Rees’s minirhizotron data at OBS, OJP, HJP94, and OA 6) Extra soil temperature data at OBS – originally logged by University of Edinburgh during BOREAS. 7) SRC’s Northern OBS climate data. 8) CS615 Volumetric Water Content (soil moisture data) 2. Investigator(s) 2.1 Principal Investigator(s) Name and Title [Identify the Principal Investigator(s) for this data set, including general affiliation if applicable] Alan Barr (PI) Climate Resarch Branch Environment Canada National Hydrology Research Centre 11 Innovation Boulevard Saskatoon, SK S7N 3H5 Canada Alan.Barr@ec.gc.ca (306)975-4324 2.2 Title of Investigation [Official name of group taken from the Fluxnet-Canada Experiment Plan.] ***To be written*** 2.3 Contact Information [Identify and give full coordinates of the person(s) most knowledgeable about the actual collection and processing of the data sets. In many cases this will be a person (or persons), other than the Principal Investigator.] Erin Thompson BERMS Coordinator Climate Resarch Branch Environment Canada National Hydrology Research Centre 11 Innovation Boulevard Saskatoon, SK S7N 3H5 Canada Erin.Thompson@ec.gc.ca (306)975-4215 Charmaine Hrynkiw Climate Resarch Branch Environment Canada National Hydrology Research Centre 11 Innovation Boulevard Saskatoon, SK S7N 3H5 Canada Charmaine.Hrynkiw@ec.gc.ca (306)975-5627 2.4 Field and/or laboratory staff: Bruce Cole, Dell Bayne, Joe Eley, Natasha Neumann, Craig Smith, Erin Thompson, Steve Enns. 2.5 Acknowledgements : Students: Jodi Axelson, Andrea Eccleston, Matt Regier, Jenny Hill, Courtney Campbell, Lisa Christmas, Kim Kovacs, Justin Beckers, Brett Reynolds. --------------------------------------------------------------------------- 3. Theory of Measurements [Theoretical basis for the way in which the measurements were made (e.g. special procedures, characteristics of the instrument, etc.).] Meteorological measurements were taken in order to understand the general climate of the Canadian boreal region and provide supporting measurements to compute carbon, water and energy fluxes. Since many different instruments were used to measure meteorological data, please refer to the user manuals referenced in the reference section for details on operation. See section on Equipment below for summarized information. --------------------------------------------------------------------------- 4. Equipment 4.1.1 Sensor/Instrument Description, Manufacturer, Calibration, Specs [This section provides a listing of the instrumentation and the characteristics of the instrumentation.] Automated stations were set up to log data from meteorological sensors. These data were either downloaded remotely via modem or stored on modules until they could be downloaded to a computer. Various types of instruments were used to measure the following variables A) temperature and humidity, B) precipitation, C) radiation, D) water table height, E) soil heat flux, F) air pressure, G) wind speed and direction, H) soil moisture and I) Campbell Scientific data loggers. A)TEMPERATURE and HUMIDITY – HMP35CF Temp/Humidity Probe, HMP45C Temp/Humidity Probe, Chromel-Constantan Fine-wire Thermocouples, Copper-Constantan Thermocouples, 107 Temperature Probe, Platinum resistance thermometer (description, manufacturer, calibration, specs, frequency of calibration, other calibration information) i) HMP35CF Temperature/Humidity Probe - Description: “The model HMP35CF probe contains a Vaisala capacity relative humidity sensor and the YSI 44002A thermistor. The probe is designed to be housed in the 41002-12 Plate Gill Radiation Shield or equivalent.” (Campbell Scientific, 1992). - Manufacturer: Campbell Scientific - Humidity Calibration: Humidity has been calibrated over a range of approximately 15%-85%. Humidity was generated in controlled lab conditions; the measurements were checked against a Hygro M3 Dew Point hygrometer. Corrections determined from these calibrations have not been applied to humidity so far. -Temperature Calibration: none. Temperature has been subject to tests, in which 3 or more instruments were compared to each other outdoors, installed on the same platform. - Thermistor Specs: Vaisala HMP35CF temperature and relative humidity probe, YSI 44002A thermistor. Resolution: 0.1 deg C. Accuracy: +/- 0.4 deg C over the range of –53 deg C to +48 deg C (overall accuracy is better than +/-0.2 deg C). The bridge resistors are 0.1% tolerance with a 10ppm temperature coefficient. - RH Specs: Resolution: 1%. Accuracy: at 20 deg C against factory references is +/-1%, against field references +/-3%, temperature dependence is +/-0.04%RH/deg C. Long term stability is better than 1% per year. Response time is 15 seconds with membrane filter. Operating temperature is –20 to +60 deg C. - Frequency of humidity calibration: once every 2-5 years. Also, comparisons between sensors installed outdoors on the same platform took place every 1-2 years. - Frequency of temperature calibration: n/a. Also, comparisons between sensors installed outdoors on the same platform took place every 1-2 years. - Other temperature/humidity calibration information: n/a. ii) HMP45C Temperature/Humidity Probe - Description: This probe contains a Platinum Resistance Temperature detector (PRT) and a Vaisala HUMICAP 180 capacitive relative humidity sensor. - Manufacturer: Campbell Scientific - Calibration: Factory calibrated. Tested by MSC Saskatoon prior to deployment in the field. - Specs (PRT): 1000 Ohms PRT, IEC 751 1/3 Class B; Temp Measurement Range: -40 to +50 deg C; Temp Output signal Range: 0.008 to 1V; Temp Accuracy: +/- 0.2 to 0.5 deg C - Specs (Humicap): RH Measurement Range: 0 to 100% non-condensing; RH Output Signal Range: 0.008 to 1 VDC; Accuracy at 20 deg C: +/-2% RH (0 to 90% RH) and +/-3% RH (90 to 100% RH); Temperature Dependence of Relative Humidity Measurement: +/-0.05% RH/degC; Typical Long Term Stability: Better than 1% RH per year; Response Time (at 20 deg C, 90% response): 15 seconds with membrance filter. - Frequency of Calibration: Factory calibrated prior to purchase only. iii) Chromel-Constantan Fine-wire Thermocouples: - Description: Chromel-Constantan thermocouples had 0.001 inch or 0.003 inch thick bare wires with welded junctions. The lead wires were soldered to 30 gauge insulated wire. The 30 gauge wire and metal shielded wire were joined via miniature connectors. - Manufacturer: Parts ordered from Omega, and instrument constructed at MSC Saskatoon. - Calibration: none - Specs: Chromel-Constantan thermocouple, 0.001” and 0.003” wire thickness. Resolution: 0.001 deg C. Accuracy: ? - Frequency of Calibration: n/a. - Other Calibration information: n/a. iv) Copper-Constantan Thermocouples: - Description: Copper-Constantan thermocouples were used for measuring soil temperature and air temperature. The soil thermocouples were 105T welded junction thermocouple probes. The air thermocouples were constructed by soldering Copper-Constantan thermocouple wire junctions, and insulating the junction with silicone. - Manufacturer (Soil thermocouples): Campbell Scientific. - Manufacturer (Air thermocouples): wire manufactured by Omega? 30 AWG. - Manufacturer (Snow Temp profile thermocouples): Thermocouple junction constructed at Queen's University, Kingston, Ontario? - Calibration: n/a. - Specs: Campbell Scientific 105T welded junction thermocouple probe. Range: -78 deg C to 100 deg C. Accuracy 0.13 deg F. - Frequency of Calibration: n/a. - Other Calibration information: n/a. v) 107 Temperature Probe - Description: 107 Temperature probe contains a Fenwal Electonics UUT51J1 thermistor (Campbell Scientific, 1983). - Manufacturer: Campbell Scientific - Calibration: factory calibration prior to purchase? - Specs: 107 Temperature probe resolution: 0.001 deg C. Accuracy: +/-0.4 deg C over the range of –33 to 48 deg C (overall accuracy is better than +/-0.2 deg C. The bridge resistors are 0.1% tolerance with a 10ppm temperature coefficient. (Campbell Scientific, 1983) - Frequency of Calibration: none since purchase? vi) Platinum resistance thermometer. - Description: - Manufacturer: - Calibration: - Specs: - Frequency of Calibration: B)PRECIPITATION - Belfort Model 5915, Belfort Model 3000, Tipping Bucket Rain Gauge, CS700-L Rain Gauge, Ultrasonic Depth Gauge Snow Depth Sensor, i) Belfort Model 5915 (Universal) Weighing Gauge - Description: Belfort precipitation gauge series 5915 are weighing-type gauges in which a weighing mechanism converts the weight of the rainfall caught by a bucket (through a circular, horizontal 8” (203mm) opening at the top of the gauge) into a resistance. (Belfort Instrument Company, 1986). The capacity of this gauge is approximately 135mm of precipitation. - Manufacturer: Belfort. - Calibration: Calibrated by MSC Saskatoon in controlled lab conditions. - Specs: Belfort 5915 accuracy: 1/3 to 1/2 of 1% F.S. Sensitivity: 0.01" (.254mm). Collector Diameter: 8" (203mm). - Frequency of Calibration: Calibrated prior to deployment in field, or when measurements were suspect. ii) Belfort Model 3000 Weighing Gauge - Description: Weighing-type gauges in which a weighing mechanism converts the weight of the rainfall caught by a bucket (through a circular, horizontal 8" (203mm) opening at the top of the gauge) into a resistance of the potentiometer? (Belfort Instrument Company, 1986). The capacity of this gauge is approximately 19.5" (495mm) of precipitation. - Manufacturer: Belfort - Calibration: Calibrated by MSC Saskatoon in controlled lab conditions. - Specs: Belfort 3000 accuracy: +/-0.15" (3.8mm) of rain span, +/-2.5% of net change for a 2" (50.8mm) to 4" (101.6mm) change, and +/-0.05% for a 2" (50.8mm) change. Range: 0"-19.5". Sensitivity: 0.025" (.6mm). - Frequency of Calibration: Calibrated prior to deployment in field, or when measurements were suspect. iii) Tipping Bucket Rain Gauge Model 525M - Description: Is a smaller adaptation of the standard Weather Bureau Tipping Bucket Rain Gauge. It measures rainfall at rates up to 2" per hour with an accuracy of +/- 1%. Output is a switch closure for each bucket tip. A tip occurs with each .1mm of rain (Campbell Scientific, 1996). - Manufacturer: Texas Electronics Inc. - Calibration: Factory Calibrated. - Specs: Range: infinite increments of .1mm; Accuracy: 1% at 2" per hour or less; Signal output: momentary switch closure activated by the tipping bucket mechanism. Switch closure is approximately 135ms. - Frequency of Calibration: Tested in controlled lab conditions prior to deployment in field or if measurements were suspect. Necessary adjustments applied to increase or decrease the amount of tips. Not officially re-calibrated. iv) Tipping Bucket Rain Gauge Model CS700-L - Description: Is a smaller adaptation of the standard Weather Bureau Tipping Bucket Rain Gauge. Its measurement range is from 0 to 500mm/hr with an accuracy better than + 2%(@ 100mm/hr). When 0.2mm of rainfall are collected the tipping bucket assembly tips and activates a reed switch. The switch closure is recorded by the datalogger pulse channel. (Campbell Scientific, 1996). - Manufacturer: Hydrological Services Pty. Ltd. (model TB-3) - Calibration: Factory Calibrated. - Specs: Range: infinite increments of 0.2mm; Accuracy: + 2% at 100mm per hour; Signal output: momentary switch closure activated by the tipping bucket mechanism. - Frequency of Calibration: Tested in controlled lab conditions prior to deployment in field or if measurements were suspect. Necessary adjustments applied to increase or decrease the amount of tips. Not officially re-calibrated. v) Ultrasonic Depth Gauge Snow Depth Sensor - Description: The UDG01 is used to find the distance from the sensor to a surface and is typically used to measure snow depth. The UDG01 operates by sending out an ultrasonic pulse and determines the time for the echo to return. - Manufacturer: Campbell Scientific - Calibration: Factory Calibrated. - Specs: Measurement Range: 0.6 to 10m; Accuracy: +/-1cm or 0.4% of Distance to the Target (whichever is greatest); Resolution: 0.5mm; Beam Acceptance Angle: Approx. 20 deg; Operating Temperature: -25 to 50 degC Standard; Operating Humidity: 5 to 95% non-condensing; Max Cable Length: 1000 feet. - Frequency of Calibration: Factory calibrated once and tested at MSC Saskatoon prior to deployment in field. Distances from sensor to target occasionally checked in the field. vi) SR50 Snow Depth Sensor - This sensor measures the distance from the sensor to a target. The SR50 determines the distance to a target by sending out ultrasonic pulses and listening for the returning echoes that are reflected from the target. The time from transmissions to return of an echo is the basis for obtaining the distance measurement. Since the speed of sound in air varies with temperature, an independent temperature measurement is required to compensate the distance reading. A simple calculation is applied to initial reading simple calculation is applied to initial readings for this purpose. - Manufacturer: Campbell Scientific - Calibration: Factory calibrated. - Specs: Range: 0.5 to 10m; Accuracy: +/- 1cm or 0.4% of distance to target (whichever is greatest) requires external temperature compensation; Resolution: 0.1mm; Beam Acceptance Angle: Approx 22 deg; Operating Temperature: -30 to +50 deg C Standard; Dimensions: Length 31cm Diameter 7.5cm; Weight: 1.3kg. - Frequency of Calibration: Factory calibrated once and tested at MSC Saskatoon prior to deployment in the field. Distances from sensor to target occasionally checked in the field. C)RADIATION i) Middleton CNR-1 Net Radiometer - Description: The Middleton CNR-1 is a pyrradiometer for the measurement of net total radiation flux (solar, terrestrial, and atmospheric). It can be used for portable or stationary applications (Carter-Scott Design, 1995) - Manufacturer: Carter-Scott Design - Calibration: Factory Calibrated. - Specs: Sensitivity: 35 to 45 microvolts/Watts per m sq; Impedance: 70-80 Ohms; Response Time: 45s @ 95%; Non-linearity: <+/-1% at 500 W per m sq; Operating Temperature: -40 to +60 deg C; Cosine Response: 4% of ideal; Spectral Range: 0.3 to 60 micro m. - Frequency of Calibration: Factory calibrated once prior to deployment in field and then calibrated at NARC in 1996?. ii) Li-Cor LI190 PAR Sensor - Description: LI-COR quantum sensors measure photosynthetically active radiation (PAR) in the 400 to 700 nm waveband. The unit of measurement is micromoles per s per sq m. The quantum sensor is designed to measure PAR received on a plane surface. A silicon photodiode with a an enhanced response in the visible wavelengths is used as the sensor (LI-COR, 1991). - Manufacturer: LI-COR - Calibration: Factory Calibrated. - Specs: Absolute Calibration: +/-5% traceable to the U.S. National Institute of Standars and Technology; Sensitivity: 8 microA per 1000 micromol per s per m sq; Linearity: Max deviation of 1% up to 10,000 micromol per s per m sq; Stability: <+/-2% change over a 1yr period; Response Time: 10 micro s; Temperature Dependence: +/-0.15% per deg C maximum; Cosine Correction: Cosine corrected up to 80deg angle of incidence; Azimuth: <+/-1% error over 360deg at 45deg elevation; Tilt: No error induced from orientation. - Frequency of Calibration: Factory calibrated once prior to deployment in field. iii) Kipp & Zonen CM11 Pyranometer (Shortwave) - Description: The pyranometer CM 11 is designed for measuring the irradiance (radiant-flux, Watts per m sq) on a plane surface, which results from the direct solar radiation and from the diffuse radiation incident from the hemisphere above (Kipp & Zonen). - Manufacturer: Kipp & Zonen. - Calibration: Factory Calibrated. - Specs: Response Time: < 15 s; Non-stability % Change Reponsitivity Per Year: +/-0.5%; Non-linearity: +/-0.6%; Directional Response for Beam Radiation: +/-10Wm2; Spectral Selectivity: +/-2%; Temperature Response: +/-1%; Tilt Response: +/-0.25%; Irradiance: 0 to 1400 W per m sq; Spectral Range: 335 to 2200nm (95% points); Sensitivity: between 4 and 6 micro V per W per m sq. - Frequency of Calibration: Factory Calibrated once prior to deployment in field. iv) Eppley PSP Pyranometer (Shortwave) - Description: - Manufacturer: EPLAB - Calibration: Factory Calibrated. - Specs: - Frequency of Calibration: Factory calibrated once prior to deployment in field. v) Eppley PIR Pyrgeometer (Longwave) - Description: The Precision Infrared Radiometer, pyrgeometer, is an instrument designed for the measurement of (unidirectional) global incoming or outgoing long-wave terrestrial radiation (EPLAB). - Manufacturer: EPLAB - Calibration: Factory Calibrated. - Specs: Sensitivity: 4 microvolts/watt per m sq approx.; Impedance: 700 Ohms approx.; Temperature Dependence: +/-2%, -20 to 40 degC (nominal); Linearity: +/-1%, 0 to 700 Watts per m sq; Response Time: 2s (i/e signal); Cosine Response: better than 5% from normalization, insignificant for a diffuse source; Orientation: No effect on instrument performance; Mechanical Vibration: Capable of withstanding up to 20g’s; Calibration: Blackbody reference. - Frequency of Calibration: Factory calibrated once prior to deployment in field. D)WATER TABLE HEIGHT i) Druck PTX Depth Pressure Transmitter - Description: - Manufacturer: Druck Incorporated - Calibration: Factory Calibrated. - Specs: Non-linearity and hysteresis: +/-0.25% BSL; Temperature error band: +/-0.3%; Compensated temp. range: -1 to 30 deg C. - Frequency of Calibration: Factory calibrated once prior to installation. E)SOIL HEAT FLUX i) Middleton CN3 Heat Flux Plate - Description: Heat transfer occurs by conduction, convection, or radiation. The CN3 Heat Flux Plate is designed to directly measure the conductive heat transmission in the medium in which the sensor is embedded. It is small and thin to offer minimal disturbance to the heat flow pattern. A temperature difference between the top and bottom faces of the CN3 generates a DC voltage from the sensor thermopile. The temperature difference is proportional to the heat flow in the medium, and the polarity of the output voltage changes in accordance with the direction of the heat flow (Carter-Scott Design). - Manufacturer: Carter-Scott Design - Calibration: Factory calibrated. - Specs: Sensitivity: 21 microV/W.m2 (typical) in fine dry sand; Impedance: 23 Ohms (typical); Temperature Range: -20 to +70 deg C; Thermal Conductivity: .4W/m.deg C; Temperature Error: .2%/deg C; Response Time: 30sec. To 95% (in air); Sensor Thermopile: copper-constantan (250 junctions). - Frequency of Calibration: Factory calibrated once prior to installation. F)AIR PRESSURE i) Setra - Description: The SBP270 Barometric Pressure Sensor is a high accuracy barometer designed for use with the CR7, 21X and CR10 dataloggers. The Setra includes a variable capacitance barometer, interface circuitry, 5 foot cable and a rain-tight enclosure. - Manufacturer: Setra, distributed by Campbell Scientific. - Calibration: Factory calibrated prior to deployment. - Specs: Pressure Range: 800-1100 mb, Accuracy: +/- 0.2mb, Long Term Stability: < +/-0.1% FS over 6mo 70 deg F. - Frequency of Calibration: once prior to deployment. G)WIND SPEED AND DIRECTION i) RMY Gill Cup Wheel Anemometer (Speed only) - Description: Utilizes a d.c. tachometer generator whose analog output voltage is directly proportional to wind speed. The cup wheel assembly has three hemisphere shaped cups formed of polypropylene which exhibit a distance constant of approximately 2.7m. - Manufacturer: R.M. Young. - Calibration: Calibration is easily checked by removing the cup wheel and driving the shaft at a known rpm by means of a synchronous motor calibrating unit or similar device. - Specs: Threshold is 0.35-0.45 m/s. - Frequency of Calibration: Once prior to deployment. ii) RMY Propeller Anemometer Model 05103 (Speed and Dir) - Description: Is a 4-blade helicoid propeller. Propeller rotation produces an AC sine wave voltage signal with frequency directly proportional to wind speed. The wind direction sensor is a lightweight vane with a sufficiently low aspect ratio. - Manufacturer: R.M. Young - Calibration: Factory Calibrated - Specs: Range: 0-60 m/s and 360 deg, Accuracy: +/-0.3m/s and +/-3deg - Frequency of Calibration: once prior to deployment. H)SOIL MOISTURE i) Campbell Scientific CS615 soil moisture probes - Description: Consists of 2 stainless steel rods connected to a printed circuit board. A shielded four-conductor cable is connected to the circuit board to supply power, enable the probe, and monitor the pulse output. - Manufacturer: Campbell Scientific - Calibration: Lab calibrated in Saskatoon prior to deployment. - Specs: +/-2% when using calibration for a specific soil. The accuracy depends on soil texture and mineral composition. - Frequency of Calibration: once prior to deployment. ii) ESI Environmental Time Domain Reflectometry Probes - Description: measures the propagation time of a signal as it travels through a transmission line (a probe) embedded in the soil. The longer the propagation time, the higher the moisture content. Propagation times, together with the probe length, soil and probe coefficients, can be converted directly into soil % moisture content. - Manufacturer: E.S.I. Environmental Sensors, Inc. - Calibration: n/a - Specs: n/a - Frequency of Calibration: n/a I) CAMPBELL SCIENTIFIC DATA LOGGERS - Description: provides the means to log information from sensors whose leads are connected to I/O cards. Control module functions include real-time task initiation, measurement processing, data storage, telecommunications and keyboard/display interaction. - Manufacturer: Campbell Scientific - Calibration: n/a - Specs: n/a - Frequency of Calibration: 4.1.2 Manufacturer of Sensor/Instrument [Name, address, and telephone number of the company that produced the instrument. If the measuring device was built by the investigator, or specially customized, please specify.] See section above. 4.1.3 Principles of Operation [Fundamental scientific basis for the way the instrument operates. This is a summary; where a full development is required, it should be placed in section 3.] See section above. 4.1.4 Source/Platform [What the instrument(s) is(are) mounted on, e.g. tower, hand held, aircraft.] - Most of the meteorological instruments were installed on double scaffold towers with stairs within. These towers were equipped with AC power and had heated huts nearby, also with AC power. Parameters measured on these towers included: air temperature, humidity, radiation and wind. - Precipitation sensors were accompanied by wind and air temperature instruments installed on a wooden platform approximately 2-3m high, 1m wide and 2m long. - Soil temperature, soil moisture, soil heat flux, water table height were measured below ground level, near the tower/hut. - Dataloggers were housed in the above mentioned heated huts or in datalogger enclosures. For more details see section on Data Description 4.1.5 Sensor/Instrument Measurement Geometry [Describe the sensor location(s), orientation, and any other parameters that affect the collection or analysis of data, e.g. field of view, optical characteristics, height, etc.] - Most above ground instruments, such as air temperature, wind and radiation sensors were installed on walk-up towers, ranging in height from 4 to 38m agl. The exceptions were near ground measurements, below 2m. These sensors were installed near the towers, on their own platforms. All sensors were installed to optimize measurement requirements. -Precipitation gauges and wind and temperature sensors associated with these gauges were installed on a platforms separate from the walk-up towers, at about 2 to 4m agl. This platform was located in a clearing approximately 50m wide. -Atmospheric pressure sensors were installed in the Huts, at approximately 2m agl. -Soil moisture, ground heat flux and soil temperature sensors were installed below the ground, beneath organic layers of leaf litter. See the section on Data Description for more information. 4.1.6 Collection Environment [Under what environmental conditions were the data collected and the instrumentation operated. This includes descriptions of the types of sites visited and factors that may effect the measurements such as temperature range experienced during data collection.] Temperatures ranged from a maximum of 35 deg C in the summer to a minimum of -40 deg C in the winter. On average, the region has a frost-free season of between 80 to 100 days (Phillips, 1990). Precipitation events in the summer ranged from trace amounts to approximately 35mm. Summer storms consisted of wind gusts up to 15 m per s, with rain, lightning and/or hail. In the winter it was common to have periods of snow, ice and frost. The average maximum snow depth for the region is between 50 and 70cm (Gray, 1981). Most of the instrumentation was exposed to these elements, with the exception of dataloggers and the instrumentation installed below the ground. 4.2 Calibration [Describe how the measurements made by the device(s) are calibrated with known standards. Specific details should be given in the subsections below.] See section on Equipment above. 4.2.1 Specifications [Record any specifications that affect the calibration of the device, its operations, or the analysis of the data collected with it.] See section on Sensor/Instrument Description above. 4.2.1.1 Tolerance [Describe the acceptable range of inputs and the precision of the output values.] See section on Sensor/Instrument Description above. 4.2.2 Frequency of Calibration [Indicate how often the instrument is measured against a standard. Also indicate any other routine procedures required to maintain calibration or detect miscalibrations. Describe also the actual practice with this device.] See section on Sensor/Instrument Description above. 4.2.3 Other Calibration Information [Give factory calibration coefficients, information about independent calibrations, history of modifications, etc.] See section on Sensor/Instrument Description above. --------------------------------------------------------------------------- 5 Site description [Standard site description, should include site location in a well referenced coordinate system, site elevation, vegetation type, etc...] - Location: Near White Swan Lake (1.6km in from route 913, 19.2km N of route 120). Lat: 53.98717° N, Long: -105.11779° W (BOREAS coordinates), elev: 628.94m - Topography: Flat, spruce bog. - Predominant vegetation: black spruce trees and moss with labrador tea as ground cover. Secondary vegetation: jack pine and tamarack, as well as cranberry. The black spruce trees ranged in height from 0-10m, and jack pine and tamarack were approximately 10-16m high. - Soil properties: Primarily sandy loam/loam till but also peat with poor drainage. The organic layer is 5-20cm deep. - Other features: This type of forest is quite boggy, and has many small pockets of standing water during wet conditions. Old Black Spruce is important as a research site because the type of forest it accounts for over 70% of boreal forest cover in Canada. 6. Data Acquisition Methods [Describe the procedures for acquiring this data in sufficient detail so that someone else with similar equipment could duplicate your measurements. Should be sufficient to include in the Methods section of a paper] 6.1 Methods of data acquisition [Describe the procedures for acquiring this data in sufficient detail so that someone else with similar equipment could duplicate your measurements. Should be sufficient to include in the Methods section of a paper] Instrumentation connected to Campbell Scientific data loggers were sampled and stored on the loggers. Most of the data were downloaded daily from dataloggers through dial-up modems using Campbell Scientific software. In cases where a datalogger was not connected to a phone line, data was downloaded either by a lap-top computer or storage module. 6.2 Sampling 6.2.1 Spatial Coverage/Geographic Location [Give enough information to locate the measurement site with suitable precision. This may be a list of sites visited, or a geographic range in the case of aircraft measurements or satellite imagery, or plot coordinates in UTM, including a clear reference to the coordinate system.] Instrumentation was installed within 50m of the main tower location. - Location: Near White Swan Lake (1.6km in from route 913, 19.2km N of route 120). Lat: 53.987 deg N, Long: -105.117 W, elev: 628.94m For more details on exact location of instruments, see section on Data Description. 6.2.2 Spatial sampling [Includes a description of spatial sampling: how many sites/samples; how were they selected, the coordinates (e.g. UTM) of the plots, how many replicates over space, etc...] See section on Data Description for more information. 6.2.3 Temporal coverage [The period(s) of time during which data was collected more or less continuously.] Data was continuously collected all year round. 6.2.4 Temporal sampling [Includes a description of the temporal sampling scheme: when was the sampling carried out (time of day), at what frequency were the measurements taken, how long did the measurements take, etc...] Most meteorological variables were sampled every 5seconds and averaged over 30min intervals. The exception to this would be snow depth and cumulative precipitation. --------------------------------------------------------------------------- 7 . Observations 7.1 Procedural Notes [Use this section to record observations made during actual data collection, which could bear on the analysis of the data, e.g. condition of site, peculiar procedures or operations, the presence of U.F.O.'s or bears, oddities in equipment function, etc.] n/a 7.2 Field Notes [If a large amount field notes exist, a reference to a separate file will be adequate.] See FieldNotes.txt --------------------------------------------------------------------------- 8 . Data Description [This section describes the data in the data set: what the data are, units, format, data characteristics.] 8.1 Data Organization [Describe how your data is organized, e.g.: by site and/or month.] Data are organized by site and month. Three types of files are available: i)Ancillary: a subset of extra Meteorological variables that were used to derive and quality assure some variables in the Main directory. These variables would not normally be used by most people, but are valuable enough to keep in the archive. ii)Main: contains the most complete sub set of meteorological variables where most have been corrected and some have been derived, but none have been gap-filled. iii)Summarized: contains data that have been summarized from the Main subset. Variables have been aggregated (eg. 4 tower top temperature measurements have become one) and gaps have been filled. Most variables are not available yet (as of Mar/04). 8.2 Image andData Format [Specify the format that the image or the data is (are) provided in.] Data are comma delimited ASCII format. 8.3 Numerical Data Characteristics [Describe the types of data submitted. On separate lines, indicate each column number followed by its header, the variable description, the unit of measurement or format of presentation, the data source or sensor, and the variable range.] Note that there are three types of Meteorology files available: 1) Ancillary 2) Main and 3) Summarized. All three are included below. This information may also be found in the "readme.txt" files found under each file type's directory. 1) Ancillary Variable List: ------------- 1 DataType="Met1" (n/a) Includes "raw" variables used to compute some other variables (corrections applied but gaps not filled.) 2 Site="SK-OBS" (n/a) Saskatchewan Old Black Spruce site 3 SubSite="FlxTwr" (n/a) Flux Tower sub-site. 4 Year (n/a) 4 digit year in UTC. 5 Day (n/a) Day of Year in UTC. 6 End_Time (UTC) End of 30min time period, in hours and minutes UTC. 7 CompDownLong (W/m2) "Compensated" downwelling longwave radiation waveRad_Abv on top of walk-up tower, mounted on a railing Cnpy_25m facing southward 15m above the canopy at 25m agl. Eppley PIR. 8 DownLongwave (W/m2) Raw thermopile output from the downwelling _Thermopile longwave sensor above. 9 DownLongwave (kOhm) Instrument body temperature of the downwelling _BodyTemp longwave sensor above. 10 DownLongwave (kOhm) Instrument dome temperature of the downwelling _DomeTemp longwave sensor above. 11 CompUpLong (W/m2) "Compensated" upwelling longwave radiation waveRad_Abv on an extended boom attached to the walk-up Cnpy_20m tower, facing southward 10m above the canopy at 20m agl. Eppley PIR. 12 UpLongwave (W/m2) Raw thermopile output from the upwelling _Thermopile longwave sensor above. 13 UpLongwave_ (kOhm) Instrument body temperature of the upwelling BodyTemp longwave sensor above. 14 UpLongwave_ (kOhm) Instrument dome temperature of the upwelling DomeTemp longwave sensor above. 15 WindSpd_At (m/s) Wind speed at Belfort weighing precip gauge. Belfort_4m RM Young propeller anemometer. 16 AirTemp_At (degC) Air temperature at snow depth sensor in SnowD_Clrg clearing. CSI 107 temp probe. 17 AirTemp_At (degC) Air temperature at snow depth sensor within SnowD_Cnpy canopy. CSI 107 temp probe. 18 AirTemp_At (degC) Air temperature at snow depth sensor located SnowD_NE near NE soil pit. CSI 107 temp probe. soilpit 19 AirTemp_At (degC) Air temperature at snow depth sensor located SnowD_NW near NW soil pit. CSI 107 temp probe. soilpit 20 MetLogger (degC) Internal Main Meteorological data logger Temp_Card1 temperature on Card1 (note - Card2 temp n/a). 21 SoilLogger (degC) Internal Soil Properties' (temp and moisture) Temp logger temperature. 22 SnowRef_Temp (degC) AM25T Multiplexer temperature. Used as a _Mplexer a reference temperature for snow temperature thermocouples. 23 MetLogger (Volts) Main Meteorological data logger battery Battery voltage. 24 SoilLogger (Volts) Soil Properties' (temp and moisture) data Battery logger battery voltage. 25 SnowLogger (Volts) Snow Propterites' (depth and temp) data Battery logger battery voltage. 26 Certification (n/a) CPI: checked by PI; PRE: preliminary. Code 27 RevisionDate (dymo Date data last revised by PI. year) 2) Main 1 DataType="Met2" (n/a) Second subset of meteorological data corrections applied but gaps not filled. 2 Site="SK-OBS" (n/a) Saskatchewan Old Black Spruce site 3 SubSite="FlxTwr" (n/a) Flux Tower sub-site. 4 Year (n/a) 4 digit year UTC. 5 Day (n/a) Day of Year UTC. 6 End_Time (UTC) End of 30min time period, in hours and minutes UTC. 7 FourWay_Net (W/m2) Derived by: (downwelling shortwave Rad_AbvCnpy - upwelling shortwave)+(downwelling longwave - upwelling longwave). 8 Middleton_ (W/m2) Net radiation; on an extended boom NetRad_ attached to the walk-up tower, AbvCnpy facing south 10m above the canopy at 20m agl. Adjusted for short and longwave sensitivities. Middleton CNR-1. 9 GlobalShort (W/m2) Global Shortwave Radiation; on top waveRad_Abv of walk-up tower, mounted on a Cnpy_25m railing facing southward 15m above the canopy at 25m agl. Kipp & Zonen CM11. 10 UpShortwaveRad (W/m2) Upwelling Shortwave Radiation; on an AbvCnpy_20m extended boom attached to the walk-up tower, facing southward 10m above the canopy at 20m agl. Kipp & Zonen CM11. 11 DownLongwave (W/m2) Derived from thermopile, body and dome _Rad_AbvCnpy_ temperatures measured by the 25m downwelling longwave sensor. Installed on top of walk-up tower, mounted on railing facing southward 15m above canopy at 25m agl. Eppley PIR. 12 UpLongwave (W/m2) Derived from thermopile, body and dome Rad_AbvCnpy temperatures measured by the upwelling _20m longwave sensor. Installed on an extended boom attached to walk-up tower, facing southward 10m above canopy at 20m agl. Eppley PIR. 13 LI_Down (umol/ Downwelling PAR Radiation; on top of PAR_Abv m2/s) walk-up tower, mounted on a railing Cnpy_25m facing southward 15m above the canopy at 25m. Li-Cor LI190. 14 LI_UpPAR_ (umol/ Upwelling PAR Radiation; on an extended AbvCnpy_ m2/s) boom attached to the walk-up tower, 20m facing southward 10m above the canopy at 20m agl. Li-Cor LI190. 15 LI_DownPAR (umol/ Downwelling PAR Radiation Rep1; below _BlwCnpy_ m2/s) canopy 18m E of walk-up tower at 1.5m 1m_#1 agl. Li-Cor LI190. 16 LI_DownPAR (umol/ Downwelling PAR Radiation Rep2; below _BlwCnpy_ m2/s) canopy 18m E of walk-up tower at 1.5m 1m_#2 agl. Li-Cor LI190. 17 BF3_Down (umol/ Not installed yet. PAR_Abv m2/s) Cnpy_25m 18 BF3_Diff (umol/ Not installed yet. usePAR_ m2/s) AbvCnpy_ 25m 19 Wetness_Abv (n/a) Not installed yet. Cnpy_25m 20 Tc_AirTemp_ (degC) Thermocouple Air Temperature; top of AbvCnpy_25m the walk-up tower, 15m above the canopy at 25m agl. In house-made fine wire unshielded Chromel-Constantan thermocouple, 0.003" thickness. 21 MetOne_Air (degC) Ventilated HMP Air Temp; installed Temp_AbvCnpy near top of walk-up tower at 24m agl. _24m Vaisala HMP temp/RH sensor in a Met One ventilator. 22 MetOneTc_ (degC) Ventilated Thermocouple Air Temp; AirTemp_Abv near top of walk-up tower at 24m agl. Cnpy_24m 30gauge Copper-Constantan thermocouple in a Met One ventilator. 23 MetOnePRT (degC) Ventilated Platinum Resistance _AirTemp_ Thermometer Air Temperature; AbvCnpy_24m near top of walk-up tower at 24m Platinum Resistance Thermometer in a Met One ventilator. 24 AirTemp_ (degC) Air Temperature; on top of walk-up AbvCnpy_25m tower mounted on north railing 15m above the canopy at 25m agl. Vaisala HMP in a gill radiation shield. 25 AirTemp_ (degC) Air Temperature; on walk-up tower in Cnpy_6m canopy at 6m agl. Vaisala HMP in a gill radiation shield. 26 AirTemp_ (degC) Air Temperature; on tree south of walk-up AbvGnd_1m tower above ground at 1m agl. Vaisala HMP in a gill radiation shield. 27 MetOne_Rel (%) Ventilated HMP Relative Humidity; Hum_AbvCnpy installed near top of walk-up tower _24m at 24m agl. Vaisala HMP temp/RH sensor in a Met One ventilator. 28 RelHum_Abv (%) Relative Humidity; on top of walk-up Cnpy_25m tower, mounted on north railing at 25m agl. Vaisala HMP. 29 RelHum_ (%) Relative Humidity; on walk-up tower Cnpy_6m in canopy at 6m agl. Vaisala HMP. 30 RelHum_Abv (%) Relative Humidity; on tree south of Gnd_1m walk-up tower in canopy at 1m agl. Vaisala HMP. 31 WindSpd_Abv (m/s) Wind Speed Tower Top; mounted on a Cnpy_26m pipe extending above the top platform of the walk- up tower, 26m agl. RM Young Propeller Anemometer. 32 WindDir_ (deg) Wind Direction Tower Top; same as Wind AbvCnpy_26m SpeedTower Top. RM Young Propeller Anemometer. 33 StdDev_Wind (deg) Standard Deviation of Wind Direction Dir_AbvCnpy at Tower Top; same as Wind Speed _26m Tower Top. RM Young Propeller Anemometer. 34 SurfPress (kPa) Barometric Pressure; in Hut, 2m agl. Setra. 35 Belfort_Cum (mm) Manually quality controlled annual Prec accumulated precipitation from weighing gauge installed on the Hut's roof. Belfort or other weighing gauge. 36 TBRG_Rain (mm) Total Tipping Bucket Rainfall over the 30 min period; installed on the Hut's roof, 4m agl. TBRG 525M changed to CSI model CS700-L on May 27, 2002. 37 SnowDepth (cm) Snow depth; mounted above a wooden _Clrng ploatform on the E side of the Hut. Ultrasonic Depth Gauge. Sensor re- moved Sept 5, 2002 - shed now built on this platform. 38 SnowDepth (cm) Snow depth; Installed under canopy _Cnpy North of Hut. 39 SnowDepth (cm) Snow depth; installed on a level spot _NEsoilpit near by the NE soil moisture pit. 40 SnowDepth (cm) Snow depth; installed over NW soil _NWsoilpit moisture pit where ground is rel- atively level. 41 WaterTable (mm) Water Table Depth; mounted in well Depth 2.05m below ground in sandy silt (all brown) with top 25-30cm consisting of peat and moss, 40m N of Hut. Druck Piezometer PTX 1230. 42 MossTemp (degC) Moss Temperature; 1.5cm below the moss _NE_1cm layer in a pit NE of the Hut. Queen’s Univ-made Cu-Co thermocouple 43 SoilTemp (degC) Soil Temperature; 2cm below the moss _NE_2cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 44 SoilTemp (degC) Soil Temperature; 5cm below the moss _NE_5cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 45 SoilTemp (degC) Soil Temperature; 10cm below the moss _NE_10cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 46 SoilTemp (degC) Soil Temperature; 20cm below the moss _NE_20cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 47 SoilTemp (degC) Soil Temperature; 50cm below the moss _NE_50cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 48 SoilTemp (degC) Soil Temperature; 100cm below the moss _NE_100cm layer in a pit NE of the Hut. Campbell Scientific Copper-Constantan 105T thermocouple. 49 MossTemp (degC) Moss Temperature; 1.5cm below the moss _NW_1cm layer in a pit NW of the Hut. Queen’s Univ-made Cu-Co thermocouple. 50 SoilTemp (degC) Soil Temperature; 2cm below the moss _NW_2cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 51 SoilTemp (degC) Soil Temperature; 5cm below the moss _NW_5cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 52 SoilTemp (degC) Soil Temperature; 10cm below the moss _NW_10cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 53 SoilTemp (degC) Soil Temperature; 20cm below the moss _NW_20cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 54 SoilTemp (degC) Soil Temperature; 50cm below the moss _NW_50cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 55 SoilTemp (degC) Soil Temperature; 100cm below the moss _NW_100cm layer in a pit NW of the Hut. Queen’s Univ-made soil temperature rod. Cu-Co thermocouple. 56 SnowTemp (degC) Temp 1cm above ground on rod installed _W_1cm W of Hut and W of main tower. Cu-Co thermocouple. 57 SnowTemp (degC) Temp 2cm above ground on rod installed _W_2cm W of Hut and W of main tower. Cu-Co thermocouple. 58 SnowTemp (degC) Temp 5cm above ground on rod installed _W_5cm W of Hut and W of main tower. Cu-Co thermocouple. 59 SnowTemp (degC) Temp 10cm above ground on rod installed _W_10cm W of Hut and W of main tower. Cu-Co thermocouple. 60 SnowTemp (degC) Temp 20cm above ground on rod installed _W_20cm W of Hut and W of main tower. Cu-Co thermocouple. 61 SnowTemp (degC) Temp 30cm above ground on rod installed _W_30cm W of Hut and W of main tower. Cu-Co thermocouple. 62 SnowTemp (degC) Temp 40cm above ground on rod installed _W_40cm W of Hut and W of main tower. Cu-Co thermocouple. 63 SnowTemp (degC) Temp 50cm above ground on rod installed _W_50cm W of Hut and W of main tower. Cu-Co thermocouple. 64 SnowTemp (degC) Temp 1cm above ground on rod installed _NW_1cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 65 SnowTemp (degC) Temp 2cm above ground on rod installed _NW_2cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 66 SnowTemp (degC) Temp 5cm above ground on rod installed _NW_5cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 67 SnowTemp (degC) Temp 10cm above ground on rod installed _NW_10cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 68 SnowTemp (degC) Temp 20cm above ground on rod installed _NW_20cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 69 SnowTemp (degC) Temp 30cm above ground on rod installed _NW_30cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 70 SnowTemp (degC) Temp 40cm above ground on rod installed _NW_40cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 71 SnowTemp (degC) Temp 50cm above ground on rod installed _NW_50cm NW of Hut and next to NW soil temp/moisture pit. Cu-Co thermocouple. 72 Certification (n/a) CPI: checked by PI; PRE: preliminary. Code 73 RevisionDate (dymo Date data last revised by PI. year) 3) Summarized 1 DataType (n/a) 2 Site (n/a) 3 SubSite (n/a) 4 Year (UTC) 5 Day (UTC) 6 End_Time (UTC) 7 FourWay_NetRad_AbvCnpy (W/m2) 8 GlobalShortwaveRad_AbvCnpy_25m (W/m2) 9 LI_DownPAR_AbvCnpy_25m (umol/m2/s) 10 AirTemp_AbvCnpy (degC) 11 RelHum_AbvCnpy (%) 12 SpecificHum_AbvCnpy (g/kg) 13 WindSpd_AbvCnpy_26m (m/s) 14 SoilTemp_2cm (degC) 15 SoilTemp_5cm (degC) 16 SoilTemp_10cm (degC) 17 SoilTemp_20cm (degC) 18 SoilTemp_50cm (degC) 19 SoilTemp_100cm (degC) 20 Belfort_CumPrec (mm) 21 EventPrec (mm) 22 CertificationCode (n/a) 23 RevisionDate (dymoyear) 4) Gap Filled Meteorology 1 DataType (n/a) 2 Site (n/a) 3 SubSite (n/a) 4 Year (UTC) 5 Day (UTC) 6 End_Time (UTC) 7 GapfilledPIPref_FourWay_NetRad_AbvCnpy (W/m2) 8 GapfilledPIPref_Middleton_NetRad_AbvCnpy_20m (W/m2) 9 GapfilledPIPref_GlobalShortwaveRad_AbvCnpy_25m (W/m2) 10 GapfilledPIPref_UpShortwaveRad_AbvCnpy_20m (W/m2) 11 GapfilledPIPref_DownLongwaveRad_AbvCnpy_25m (W/m2) 12 GapfilledPIPref_UpLongwaveRad_AbvCnpy_20m (W/m2) 13 GapfilledPIPref_LI_DownPAR_AbvCnpy_25m (umol/m2/s) 14 GapfilledPIPref_LI_UpPAR_AbvCnpy_20m (umol/m2/s) 15 GapfilledPIPref_LI_DownPAR_BlwCnpy_1m_#1 (umol/m2/s) 16 GapfilledPIPref_LI_DownPAR_BlwCnpy_1m_#2 (umol/m2/s) 17 GapfilledPIPref_Tc_AirTemp_AbvCnpy_25m (degC) 18 GapfilledPIPref_MetOne_AirTemp_AbvCnpy_24m (degC) 19 GapfilledPIPref_MetOneTc_AirTemp_AbvCnpy_24m (degC) 20 GapfilledPIPref_MetOnePRT_AirTemp_AbvCnpy_24m (degC) 21 GapfilledPIPref_AirTemp_AbvCnpy_25m (degC) 22 GapfilledPIPref_AirTemp_Cnpy_6m (degC) 23 GapfilledPIPref_AirTemp_AbvGnd_1m (degC) 24 GapfilledPIPref_MetOne_RelHum_AbvCnpy_24m (%) 25 GapfilledPIPref_RelHum_AbvCnpy_25m (%) 26 GapfilledPIPref_RelHum_Cnpy_6m (%) 27 GapfilledPIPref_RelHum_AbvGnd_1m (%) 28 GapfilledPIPref_WindSpd_AbvCnpy_26m (m/s) 29 GapfilledPIPref_WindDir_AbvCnpy_26m (m/s) 30 GapfilledPIPref_SurfPress (kPa) 31 GapfilledPIPref_Belfort_CumPrec (mm) 32 GapfilledPIPref_SnowDepth_Clrng (mm) 33 GapfilledPIPref_SnowDepth_Cnpy (mm) 34 GapfilledPIPref_SnowDepth_NEsoilpit (mm) 35 GapfilledPIPref_SnowDepth_NWsoilpit (mm) 36 GapfilledPIPref_MossTemp_NE_1cm (degC) 37 GapfilledPIPref_SoilTemp_NE_2cm (degC) 38 GapfilledPIPref_SoilTemp_NE_5cm (degC) 39 GapfilledPIPref_SoilTemp_NE_10cm (degC) 40 GapfilledPIPref_SoilTemp_NE_20cm (degC) 41 GapfilledPIPref_SoilTemp_NE_50cm (degC) 42 GapfilledPIPref_SoilTemp_NE_100cm (degC) 43 GapfilledPIPref_MossTemp_NW_1cm (degC) 44 GapfilledPIPref_SoilTemp_NW_2cm (degC) 45 GapfilledPIPref_SoilTemp_NW_5cm (degC) 46 GapfilledPIPref_SoilTemp_NW_10cm (degC) 47 GapfilledPIPref_SoilTemp_NW_20cm (degC) 48 GapfilledPIPref_SoilTemp_NW_50cm (degC) 49 GapfilledPIPref_SoilTemp_NW_100cm (degC) 50 CertificationCode (n/a) 51 RevisionDate (dymoyear) 8.3.1.6 Sample Data Record 8.4 Image Data [Describe the data submitted, with subsections 7.4.1 through 7.4.13 (below) being represented as columns in a table Example: Identifier:OBS02031HH.PIX Date of Acquisition (UTC):31 January 2002 Time of Acquisition (UTC):16:13 Sensor / Mode:RADARSAT-1 SAR Standard Beam S1 Wavelength (nm) / Frequency (GHz):Standard Platform Altitude (magl):N/A Spatial Ground Resolution (m):30 Incidence Angle - Average:N/A Incidence Angle - Minimum:20.0 Incidence Angle - Maximum:27.4 Polarization:HH Gain Control:Automatic Flight Azimuth:Ascending Scene Centre:53.80206 N 104.61797 W 8.4.1 1 Image Identifier [A unique image file name that the image will be archived as, e.g. OBS02031HH.PIX.] 8.4.2 2 Date of Acquisition [As UTC.] 8.4.3 Time of Acquisition [Time as UTC; to allow later users to reproduce such things as sun angle.] 8.4.4 Sensor [Identify the imaging sensor and mode used.] 8.4.5 Wavelength [The wavelength range or frequency used. If settings are fixed, the descriptor "standard" can be used.] 8.4.6 Platform Altitude [The height of the sensor above the ground surface (m). If the altitude is fixed, such as for satellite platforms, N/A may be used.] 8.4.7 Ground Spatial Resolution [The smallest resolvable unit on the ground (m).] 8.4.8 .7 Incidence Angle - Average [The average angle from vertical.] 8.4.9 .8 Incidence Angle - Minimum [The minimum angle from vertical.] 8.4.10 Incidence Angle - Maximum [The maximum angle from vertical.] 8.4.11 Polarization [The polarization set on the sensor.] 8.4.12 Gain Control [Automatic or manual gain control.] 8.4.13 Flight Azimuth [Identify the direction of travel of the platform. For satellite-based platforms, Ascending or Descending is sufficient.] 8.4.14 Scene Centre [Give the scene centre in lat/long format.] --------------------------------------------------------------------------- 9 . Data Manipulations [This section describes the steps by which the data were processed to their final form.] 9.1 Post Processing and Calculated Variables [Specify all post-treatment of data, including data processing steps and calculations. Include relevant equations with definitions of terms and units.] Initially, data were recorded on a variety of loggers in various formats. Data from these loggers were brought into a common format and were organized by site, month and data type. Once data were organized into a common format, the first level of quality control was applied. This first stage of quality control included: - one-time "hard wired" fixes to resolve mistakes in data logger programming. - range and limit checking - application of calibration coefficients, and fixing of problems like incorrect wiring and conversion of units where required - manual exclusion of bad data where they could not be detected by automated range checking procedures - merging of manually quality controlled elements (including Atmospheric Pressure, Snow Depth, and Manual Accumulated Precipitation. - and computation of derived or adjusted elements (including, Top-of-the-Atmosphere Shortwave Radiation, Downwelling and Upwelling Longwave Radiation Above Canopy, Net Radiation Above Canopy, and Four Way Net Radiation Above Canopy. This first stage of quality control was conducted automatically as part of the automated daily data retrieval. Once a week, the data were also plotted and inspected visually, using a graphical user interface that automated the plotting and data inspection processes. After the 1st stage of quality control was completed, some additional adjustments were applied to these climate data in a second stage of QC, which included - correction of RH for a maximum value that exceeds 100% - correction of Shortwave_Rad and PAR for nighttime zero offsets - gap filling of some parameters where the estimates were judged to be highly accurate, e.g., the Longwave (PIR) data were analyzed to sort out the relationship between compensated and derived outputs, and then this relationship was used to fill gaps in the derived output. Following the 2nd stage of QC above, data gaps in some key variables were filled in a third stage of processing. Most data gaps were filled using statistical relationships with related data from the same site or the other sites. These data files are not available on the FC DIS, but are available upon request (PI or data contact). 9.2 Special Corrections/Adjustments [List any 'special' corrections/adjustments made to portions but not all of the data to make it compatible with the data set as a whole.] The following are details on quality control procedures, stage 1. a) Limit checking and Range Checking This procedure sets out-of-limit data to missing. Two types of checks are performed, these include: · Rate-of-change checking; a maximum rate of change per time increment is set (i.e. air temperature must not exceed a rate of change per 30min period of 30 deg C, otherwise it will be set to missing). These values were purposely set high, so as not to exclude anything that might be real. · Limit Checking: each variable was assigned an absolute maximum and minimum value that recorded data was required to fall in to. Example: air temperature had to fall in between 45 and -50 deg C, otherwise it would be set to missing. These limits were also set high so that anything remotely real would not be excluded. Limit checking was used to flag questionable data but the data were not automatically excluded. b) Application of Calibration Coefficients and fixing of problems like incorrect wiring as well as conversion of units. Known corrections were listed in a separate file that was queried by the 1st level QC program. See "e) Derivation of new parameters" below for more information on these corrections. c) Manual exclusion of bad data Bad data that could not be detected by automated range checking procedures were excluded manually, in a semi-automated program. Outliers that could be recognized by the human eye were manually selected using a data visualization program that enabled the user to point and click on "bad" points. Once these points were selected, a record of the "bad" element and the time at which these data occurred were automatically written to a data exclusion file. The 1st stage QC program then automatically set these manually rejected data to "missing". d) Merging of manually quality controlled elements Some elements were quality controlled manually, including snow depth, accumulated precipitation and atmospheric pressure (one time only). These elements were created in separate files that were merged into the main meteorological files at the first level of QC. Manual quality control of procedures are explained below: Atmospheric Pressure -------------------- - two pressure sensors of the same type were installed at each site, one in conjunction with the flux measurement program and the other with the meteorological program. - when the met pressure data were missing, data from the flux pressure sensor were filled in where possible. Snow Depth (non-gapfilled found in “Main” data files) ---------- The procedure for quality controlling snow depth in the Met2 files was partly manual and partly automated: - Snow depth in its raw form is recorded as the sensor height above ground, in mm. - First, the amount to subtract from the raw snow depth value to get actual snow depth was determined by manually browsing for a stable reference period of about one week before and after the snow season, where the instrument was “seeing” bare ground. - If this period was not manually found, then an automated procedure was in place do this, although typically it would not do as good a job as the human eye. o So far, it’s likely that this hasn’t happened yet. - In most cases, where the reference period did not change significantly over the snow season, one manual reference was used (usually a Fall measurement). - The snow depth was then calculated by subtracting the sensor measurement from the manual reference. - Example, the height of the sensor on Oct 31, 1998 was 2000mm and on April 15, 1999, it was 2000mm, therefore snow depth was computed by: (2000 - Distance from Sensor to Snow) = Actual Snow Depth - Ground truths, such as snow surveys and manual measurements of the height of the sensor above ground were also used to confirm/refute that the sensor was/was not working properly. Snow Depth (gapfilled – found in “MainGapfilled” data files) ---------- The procedure for quality controlling and gap filling snow depth in the Gap Filled files was as follows: - Any erratic summer data was cleaned up. - Some smoothing applied to noisy data o A filter called SGOLAY (in Matlab) was used to remove high frequency noise. The goal was to remove daily noise, which was usually the result of problems with the reference temperature required by the instrument to make its measurement. This filter works almost like a running mean. - Gaps were filled using a 5 day moving window linear regression. - If the height of the sensor changed significantly from what it was at the start of the season to what it was at the end of the season, a linear correction was applied. This correction was determined by computing the slope and intercept between the start and end times of the period in which the correction was to be applied. For example, if the height of the sensor on Oct 31 was 2000mm, but at the end of April, it was 1970mm, this type of correction was warranted. Manual snow depth measurements were also used to confirm that these linear corrections were justified. - Example: RawDepthPreSnow = Height of Sensor Above Ground on Last Snow Free Day = 1963mm RawDepthPostSnow = Height of Sensor Above Ground on 1st Snow Free Day = 2002mm TimePreSnow = Date of Last Snow Free Day = Nov 5, 2000 TimePostSnow = Date of 1st Snow Free Day = Apr 22, 2001 Slope = (TimePreSnow:RawDepthPreSnow,TimePostSnow:RawDepthPostSnow) Intercept = (TimePreSnow:RawDepthPreSnow,TimePostSnow:RawDepthPostSnow) Adjusted Snow Depth =(Slope*Time+Intercept-RawDepth) Manual Cumulative Precipitation ------------------------------- The procedure for accumulating Belfort Precipitation was a manual one. The procedure was quite simple (although a bit labour intensive): 1. If data was from a Belfort Universal (150mm) gauge, the 30 minute values were rounded off to the nearest 1/10th mm 2. If data was from a Belfort 3000 (500mm) gauge, the 30 minute values were rounded off to the nearest mm (note that the above steps may have already been done by the logger or an automated qc procedure) 3. Using the difference between 30 minute observations, a 30 minute accumulation was calculated for each 30 minute period. 4. The data was manually screened for gauge servicing (the dates of those were cross referenced with field notes when possible). 5. As a check, the difference between the last total weight before servicing and the last accumulated weight after the previous servicing was calculated. It was known that the Belfort gauges were usually accurate when it came to measuring the total weight in the bucket...this is what they do well. 6. The total of the 30 minute accumulations should have matched the difference in bucket contents as calculated above. However, it never did! 7. The observations were manually screened, one at a time, in order to distinguish between real and fictitious measurements. The most common error was rounding error computed by the datalogger. ie. A small change in the signal from the gauge may have alternated between a value of 100.64 mm (which = 100.6 mm) and 100.65 mm (which = 100.7 mm), resulting in 0.1 mm of false precip. These were systematically removed from the totals. For the most part, by eliminating these, the weights would match between servicings. Evaporation was another problem that made the weights mismatched. Eliminating this factor involved some judgment as to when evaporation occurred. It then had to be determined if the next positive weight increase was a result of the rounding problem or actual precipitation. The TBRG data helped in the summer but it was a judgment call in the winter. There were also instances where someone dumped fluid into the gauge. These false values also had to be removed. e) Derivation of new parameters. The following new elements were computed at the first stage of QC: - Top-of-the-Atmosphere Shortwave Radiation - Derived Downwelling Longwave Radiation - Derived Upwelling Longwave Radiation - Adjusted Net Radiation above canopy - Four Way Net Radiation The DerivedDownwelling and DerivedUpwelling LongwaveRadAboveCanopy from the Eppley PIR longwave radiometer were computed from the instrument’s thermopile output, body temperature and dome temperature as follows. First, for both Down- and Upwelling Longwave radiometers, BodyTemp (TK) and DomeTemp (TK) were computed from the logged thermistor resistances (kOhm): c1=1.0295e-3; c2=2.391e-4; c3=1.568e-7; TK=1/(c1+c2*ln(1e3*kOhm)+c3*(ln(1e3*kOhm)); The value for TK was refiled in place of kOhm and the units were changed from (kOhm) to (K). Second, the DerivedDownwelling and DerivedUpwelling LongwaveRadAboveCanopy were estimated from the respective PIR’s Thermopile output, BodyTemp, and DomeTemp. sigma=5.6705e-8; k=3.2; DerivedDownLongwaveRadAboveCanopy = ThermopileOutput + sigma*BodyTemp^4 - k*sigma*(DomeTemp^4-BodyTemp^4); The Middleton CN1-R Net Radiation above canopy was adjusted for differences in the instrument’s short- and longwave sensitivities, using three ‘NetRad’ adjustment parameters: c, cl, and cs, where c is the multiplier used in the data logger program, cl is the correct long-wave calibration factor, cs is the correct short-wave calibration factor and Rn on the RHS of the equation is the logged value for Rn. Rn = Rn*cl/c if Rn<=0; Rn = Rn*cs/c if Rn>0; FourWayNetRadAboveCanopy was computed as: GlobalShortwaveRadAboveCanopy - UpShortwaveRadAboveCanopy + DerivedDownLongwaveRadAboveCanopy - DerivedUpLongwaveRadAboveCanopy. 9.2 The following are details on quality control procedures, stage 2 (applicable to BERMS OA, OBS and OJP Met2 Data only). Also see "Equations" document for formatted version of the equations in this section. a) Correction of RH, as measured by Vaisala HMP35 and 45C, for maximum values exceeding 100%. Each sensor has a unique maximum range which exceeds 100%. This maximum RH is independent of temperature for temperatures above 0.0 degC and appears to be stable over time. We use this value of maximum RH for each sensor as a one point calibration to adjust RH to bring the maximum RH back down to 100% as: RHadj = RH*100/RHx (1) Note that the same value for RHx is used above and below freezing although RHx(T) falls off at temperatures below 0 degC. Vaisala has designed the HMP so that the vapour pressure is calculated as es(Ta)*RH/100% using a saturated vapour pressure es(Ta) that is with respect to water at all temperatures. There is no need to estimate es(Ta) with respect to ice at sub-zero temperature because the low temperature effect is dealt with in the change of RHx(T) with temperature at sub-zero temperatures. The fCalcRHgt100 option estimates and files the value of RHx for the entire history of HMP deployment at each site and HMP level. RHx is computed as follows. - For each HMP level and site, we first input the history of sensor deployment. - For each sensor deployment, we input the data and stratify it by temperature using 5 degC increments. - For each 5 degC increment, with more than 500 data points between 0 and 20 degC, we estimate RHx as the 99.8% percentile of the data. This percentile excludes one outlier for every 500 data points. This procedure produces up to four estimates of RHx, for 0-5, 5-10, 10-15 and 15-20 degC. - If any of the individual estimates of RHx is more than 1.5% from the mean value, it is excluded. - If two or more acceptable values remain, they are averaged to estimate RHx. b) Correction of Shortwave Radiation and PAR for nighttime zero offsets. The adjustments are usually near zero for UpShortwave_Rad. For UpShortwave_Rad, they are near zero when the ventilator fan is unplugged, +2-3 W m-2 when the ventilator fan is on and operating properly, and +5-15 W m-2 when the ventilator fan is on but stalled. The zero offset adjustments are computed from nighttime data when Shortwave_Rad is zero, assuming that the day and nighttime offsets are the same. A smooth curve is fit through the nighttime data (the zero offset) by: - calculating the mean for each night - rejecting nighttime means with too few observations, a high standard deviation, or a large mean change from one night to the next - interpolating the nightly mean offsets to each 30-min period. When the nighttime offsets are unstable (i.e., the standard deviation is high or the mean changes from one night to the next).The data from that night and the adjacent days are rejected. When the nighttime offsets are stable, the interpolated 30-min offsets are applied as adjustments both day and night. The adjustment brings the nighttime means to zero and adjusts the daytime values upwards accordingly. 9.2.3 The following are details on quality control procedures, stage 3 (applicable to BERMS OA, OBS and OJP Met3 Data). Also see "Equations" document for formatted version of the equations in this section. a) Gap filling in the “Summarized” (Met3) files is given below. This methodology will slowly be phased out in favor of that used to produce the “MainGapFilled” data files (See section below for information on the “MainGapFilled” data files). Non-statistical methods There are programs in place to fill gaps in variables that are used to derive Down- and UpLongwave_Rad from the Eppley PIR. The best estimates of Down- and UpLongwave_Rad are derived (DerivedDown- and DerivedUpLongwave_Rad) based on three logged variables: the raw PIR thermopile output (mVpile, logged as Down- and UpLongwave_Thermopile), body temperature (Tb (K)) and dome temperature (Td (K)) as: DerivedLongwave_Rad = mVpile + S* Tb^4 – c*S(Td^4-Tb^4) (2) Where mVpile is the raw longwave sensor (Eppley PIR) thermopile output in mV, S is the Stephan-Boltzmann constant, Tb is the instrument’s body temperature in K and Td is the instrument’s dome temperature in K, c is an empirical correction factor (set to 3.2), A second estimate of Longwave_Rad can be made from the output of the Eppley PIR’s compensation circuit, but is not recommended for use because of two important differences from (2): - The compensation circuit adds bias. Ideally, the compensated output (mVcompensated) should equal (mVpile+ S* Tb^4) from (2). In practice, however, the compensation circuit is not ideal; it requires an internal battery, which can and does fail, and its performance is sensitive to temperature and varies among instruments. The result is an instrument- and temperature-dependent Bias between the compensated output mVcompensated and (mVpile+ S* Tb^4): Bias = mVcompensated – (mVpile +S*Tb^4) (3) The Bias (3) is a well-behaved function of body temperature, and the Bias=f(Tb) relationship is unique for each instrument-battery pair. - The compensated output does not include the third (c*S(Td^4-Tb^)) term from (2). This fills gaps in Down- and UpLongwave_Thermopile and Tb based on the output from the PIR’s compensation circuit. Gaps are filled by fitting a third-order polynomial to the Bias=f(Tb) relationship (3) for each instrument-battery pair, and then calculating Longwave_Thermopile or Tb from (2). Note that Vcompensated is more commonly missing than Vpile, so that there are many instances when gaps in Longwave_Thermopile cannot be filled from the compensated Longwave_Rad. Statistical methods *** This section is being written and is VERY incomplete. Gap filling using statistical methods is done in many ways, depending on the type of missing data and the type of data that are available for gap filling. One-off Regression method Gaps in Tb and Td are only filled when only one of the two is missing. The Td - Tb difference is modeled by multiple linear regression with GlobalShortwave_Rad and Longwave_Thermopile as the independent variables. Missing temperature are estimated from non-missing temperatures and the modeled temperature difference. The analysis is slightly more complicated for the tower-top PIR that measures DownLongwave_Rad; this PIR is ventilated and the regression relationship changes when the ventilator fan stalls. It is thus necessary to do separate analyses for periods when the fan is stalled and not stalled. Difference interpolation methods Regression methods. (Or Statistical) Gap filling is implemented using MatLab. The gap filling options include: -Small gaps in temperature and humidity. Where data gaps are for one or two periods only and there is a proximate measurement of the same variable, missing values from A are estimated using data from nearby sensor B, as follows. The difference between the two sensors (B-A) is calculated for all periods when both are present. The difference (B-A) is then linearally interpolated to periods when A is missing, and missing values of A are estimated as B-(B-A) where (B-A) are the interpolated values. -Statistical relationships (linear regression and linear regression forced through the origin) with nearby, associated variables. Gap filling HMP data when data is missing From one sensor only and for short periods only (<=3 periods). Missing values from sensor A are estimated using data from nearby sensor B, as follows. The difference between the two sensors (B-A) is calculated for all periods when both are present. The difference (B-A) is then linearally interpolated to periods when A is missing. Lastly, missing values of A are estimated as B-(B-A) where (B-A) are the interpolated values. -This effectively fixes the HMP dropout problem during phone calls. For HMP data, gaps are filled for longer periods and where all data are missing, based on linear regression with the most proximate measurement. Done from q. 9.3 Gap filling (applicable to files found in the “MainGapfilled” directory). A summary of gap filling procedures is given here. A more detailed document is available upon request from the PI. Operationally, gaps were filled one year at a time considering all possible related and available time series, however, only the variable that was most closely related to the missing variable was used to gap fill. Various methods were used depending on the size of the gap and what variable needed to be filled. Here are the methods used: - Interpolated Difference o For this method, the difference between two similar variables was interpolated linearly. The difference between y and x were calculated for two points on either side of the missing section of data (when both were not missing). The difference was then estimated by linear interpolation for periods when y was missing and x was not. Gaps were then filled for y by adding the difference to x. - 5-Day Moving Window Linear Regression o This method used a linear regression to determine the relationship between the missing and non-missing variables. A flexible window of 240 (5d) non-missing data points, or 120 points, on either side of the missing section of data was used to determine this relationship and hence predict the missing section of data. - Moving Window Linear Regression, Same Time of Day o This method used a moving window, as in the 5-Day linear regression method above, except that the data were stratified by time of day (into 48 periods). A regression was done separately within each stratum, using a flexible window of 12 non-missing data pairs or 6 points (6 days) on either side of the section of missing data. Example: if data were missing at 1200 UTC, then a regression was calculated using data from 1200 UTC only, over a period of 6 days on either side of the missing section of data. Given below are the methods and variables used to gap fill, organized by variable type: - Global Shortwave Above Canopy o Moving Window Linear Regression Same Time of Day (forced through origin) for all gap sizes o Variables used to fill are either Global Shortwave or Downwelling PAR Above Canopy, however, if the gap couldn’t be filled by these two, then it’s possible that Upwelling Shortwave and Upwelling PAR were used. - Upwelling Shortwave Above Canopy o Moving Window Linear Regression Same Time of Day (forced through origin) for all gap sizes o Variables used to fill include Upwelling or Downwelling Shortwave or Upwelling or Downwelling PAR Above Canopy - Downwelling PAR (Above Canopy) o Moving Window Linear Regression Same Time of Day (forced through origin) o Variables used to fill are either Global Shortwave or Downwelling PAR Above Canopy, if the gap wasn’t filled by these two, then it’s possible that upwelling Shortwave and PAR were used. - Downwelling PAR (Below Canopy) o Moving Window Linear Regression Same Time of Day (forced through origin) o Any PAR variable could have been used - Upwelling PAR (Above Canopy) o Moving Window Linear Regression Same Time of Day (forced through origin) o Variables used: Downwelling or Upwelling PAR and Global Shortwave and Upwelling Shortwave (Above Canopy only for all). - Derived Downwelling Longwave Above Canopy o Interpolated Difference for small gaps (<24 periods, or 12h) and 5-Day Moving Window Linear Regression for large gaps (24 periods or more). o Variables used: Derived Downwelling Longwave, Derived Upwelling Longwave, Down or Upwelling Shortwave Above Canopy, Down or Upwelling PAR Above Canopy. - Derived Upwelling Longwave Above Canopy o Interpolated Difference for small gaps (<24 periods, or 12h) and 5-Day Moving Window Linear Regression for large gaps (24 periods or more). o Variables used: Derived Downwelling Longwave, Derived Upwelling Longwave, Down or Upwelling Shortwave Above Canopy, Down or Upwelling PAR Above Canopy. - 4-Way Computed Net Radiation Above Canopy. If any gaps existed in the gap filled data, it would have been due to one of the 4 components being missing. First, gaps were filled in the components that were missing and then the 4-Way was re-computed. - Net Radiation Above Canopy o Moving Window Linear Regression Same Time of Day o Variables Used: Net Radiation Above Canopy (also included the 4-Way Computed Net Radiation) and Downwelling or Upwelling Shortwave Above Canopy. - Air Temperature (any height, any type of measurement) o Interpolated Difference for small gaps and Moving Window Linear Regression Same Time of Day for large gaps. o Variables Used: Air Temperatures, Shallow (<=20cm depth) Soil Temperatures and Tree Temperatures (any level or location). - Shallow Soil Temperature (<=20cm depth) o Interpolated Difference for small gaps and Moving Window Linear Regression Same Time of Day for large gaps. o Variables Used: Air Temperatures, Soil Temperatures and Tree Temperatures (any level or location). - Tree Temperatures (any location or tree depth) o Interpolated Difference for small gaps and Moving Window Linear Regression Same Time of Day for large gaps. o Variables Used: Air Temperatures, Shallow (<=20cm depth) Soil Temperatures and Tree Temperatures (any level or location). - Deep Soil Temperature (50cm and 100cm depth) o Interpolated Difference for all gap sizes. o Variables Used: All Soil Temperatures (any depth). - Relative Humidity (any height) o Interpolated Difference for small gaps and Moving Window Linear Regression Same Time of Day for large gaps o Only RH used to fill gaps. - Soil Heat Flux (any location or depth) o Interpolated Difference for small gaps and Moving Window Linear Regression Same Time of Day for large gaps o Only Soil Heat Flux used to fill gaps. - Wind Speed (any location or height) o Interpolated Difference used for small gaps and 5-Day Moving Window Linear Regression for large gaps. o Only Wind Speed used to fill gaps. - Atmospheric Pressure o Interpolated Difference used for small gaps and 5-Day Moving Window Linear Regression for large gaps. o Only Atmospheric Pressure used to fill gaps. --------------------------------------------------------------------------- 10 . Errors and Limitations [This section describes an error analysis for the data.] 10.1 Sources of Error [Describe what factors of the instrument or environment may introduce errors in the observations.] See section on "Known Problems with the Data" for more information. 10.2 Quality Assessment 10.2.1 Data Validation by Source [Describe all efforts to validate the data by the submitter, e.g. comparisons with data from other investigators.] Data were quality assured by comparing simiar variables at the same site and also by comparing variables at different sites within the BERMS area. This was done weekly, in near real time, to ensure that problems were flagged and fixed in a timely manner. 10.2.2 Confidence Level/Accuracy Judgment [Subjective discussion of data quality.] The data submitted are of good quality with minimal amount of errors. However, any measurement is not perfect, and the user must be aware of the limitations of the instrumentation. 10.2.3 Measurement Error for Parameters [Quantitative error estimates.] See section on Equipment for more instrument specs. 10.2.4 Additional Quality Assessments [May include visual review of plots, etc.] 10.3 . Limitations and Representativeness [Provide warnings on the use of the data, e.g. data were collected under drought conditions relations between variables may be different when things are wet, as well as known problems. Discuss how representative your data is, eg: of the landscape, climate, footprint, etc.] See section on Known Problems below for more information. 10.4 Known Problems with the Data Problems to be aware of at all Saskatchewan BERMS sites: - Beware that there was an unusual amount of frost build-up, particularly on the radiation sensors, during the winter of 1997/98. Frost problems have occurred in other years, but not to the same extent. Frost can either cause a drop or an increase in radiation measurements depending on the sensor. Frost can also cause the propellers on wind instruments to seize up and not record wind speeds above 0 m/s. - Non-zero shortwave and PAR at night during some periods: some possible causes of this include poor wiring and/or stray voltages from the ventilation fan. - Specific problems have been listed in the file: QC_Notes.txt/xls. Searches on specific variables, study sites and time can be made in this file. To avoid most of these problems, use the Qcd version of the Met data (called *.AL1 files). - Relative humidity measured by Vaisala HMPs often drift above 100%. Nothing has been done to correct this problem at the first stage of QC. Higher level QCd files, which will be available at a later date will make adjustments to RH, particularly for values near and above 100%. - Some Vaisala HMP temperatures would often go missing for the one 30min period in which phone downloads took place. This problem remained unresolved. - Downwelling and Upwelling Longwave radiation as measured by the Eppley PIR sensor (variables: DownLongwave_Rad_AbvCnpy and UpLongwave_Rad_AbvCnpy) were OK as long as the battery in the sensor was fully charged. If this battery began to die, the measurement would slowly drop off the acceptable range. If these data are found to be missing, it was likely that the battery in the sensor was dying or dead. In this case, use DerivedDownLongwave_Rad_AbvCnpy_37m and DerivedUpLongwave_Rad_AbvCnpy_37m instead. - Some gaps in the thermocouple air temperature measurements exist due to the fine wires on the sensor breaking. Data problems at OBS: - 1997 to 1999: The occasional non-zero downwelling and upwelling shortwave data spike or small excursion from 0 occurred at night from 1997-1999. These bad data were mostly positive, but were occasionally negative. Unfortunately, the cause of these problems was unknown. - 1999: Downwelling shortwave had a zero offset during 1999, which could be detected at night. This was likely symptomatic of a problem during the daytime too. - 1998 March to May: Bad upwelling PAR excluded from QCd version of data files from March until May 1998. - 1998 May 2000 July: The Global downwelling shortwave radiometer measurements were too low due to some shadows cast on the instrument: near 2400 UTC May through October 1998; near 1600 UTC from 1 Apr 1999 to 1 Oct 1999 and; near 1600 UTC from 1 Apr, 2000 to 18 July, 2000. - 1997 October: Temperature at 1m was intermittent/bad from May to Oct 1997. - Temperature data at 1m, 6m and in the Met One Ventilator at 24m often went missing during data downloads via modem (mostly in 1999). - 1997 July to 1998 May: Tipping bucket precipitation was unavailable until July 1997 and then the sensor started going bad in August 1997. The next useable data started in May 1998. - 1998 June to October: Belfort precipitation was 15% lower than the tipping bucket precipitation from June to October, 1998. - ????: The cup wheels on the anemometer at the Belfort rain gauge sometimes froze up during the winter. This could be detected when the wind speed stayed at 0 for days at a time. - 1998: There were many gaps in the data record in early in 1998 because an electrical short in the Volumetric Water Content sensors caused the datalogger to crash. - 2000 March to October: Problems with Water Table Depth Sensor from late March, 2000 until late Oct, 2000. The exact cause of these problems was not isolated. However, it could be one of three things: 1) a voltage problem, 2) problem of water in the sensor, or 3) problem of the sensor being packed with mud. - 2000 to 2001: Snow depth sensor went bad/missing during spring for 2000 and 2001. Therefore, no snow depth data exists for the spring snowmelt period for these years. - 2000 March 27: Instruments on the walk-up tower were completely rewired between 27 March and 20 March, 2000. Note that data gaps exist during and following this period. As well, some sensors were not working properly for a period after this re-wire, those included: Snow depth sensor, upwelling longwave thermopile, wind speed/dir, PRT temperature and atmospheric pressure. Note that these bad atmospheric pressure data have been replaced by pressure data from Univ of BC's sensor. - 2000 May 3 to 30: The Belfort 3000 rain gauge was not responding well to precipitation from May 3, 2000 until the instrument was serviced on May 30, 2000. - 2000 June 9 (DOY 161) to July 18 (DOY 200): TCAir_Temp_AbvCnpy_25m missing from day 161 to 200, 2000. - 2000 June 21 (DOY 173) to July 17 (DOY 199): Wind_Spd_AtBelfort3000_4m - This wind speed has been programmed so that it should never go below 0, however it drifted below 0 from about 21 June (day 173) until 17 July at 0230. A possible clue to what may have gone on was that the power supply (measured as MetLogger_Battery and SoilLogger_Battery voltages) dropped slightly at about the same time. - 2000: MetOneAir_Temp_AbvCnpy_24m inconsistently missing 0800 and/or 1100 and/or 1230 during June when data was downloaded via modem - 2000: Air_Temp_Cnpy_6m consistently missing 0800, 1100 and/or 1230 during June to present due to communications and/or data downloads - 2000: Problems with Snow_Depth measurements spiking or missing in 2000. - 2000: Soil_Tem_NW_100cm spikes low by .1degC per 30min period several times in 2000. - 2000 August 20 to October 26: WaterTable_Depth missing from 20 Aug to 26 OCt 2000. - 2000 August 23 11:30 UTC to October 26 16:00 UTC: Surf_Press - dropped from ~940 to 800mb (missing) periodically between 23 Aug 1130 1 Oct 0430, after that it is missing. Repaired on 26 Oct 1600 by fixing a loose wire at the datalogger - 2000: Problems with DownLongwave - derived Downwelling Longwave exceed Upwelling Longwave several times in 2000. - 2001 March 31 (DOY 90) to May 7 (DOY 127): TBRG_Rain Gauge not reporting tips from day 90 2001 to day 127 2001. - 2001: TBRG_Rain gauge accumulations fall behind Belfort 3000 gauge during 2001. - 2001 December 6 (DOY 340) to Dec 17 (DOY 351): Main wire powering climate loggers was found to be loose, backup power did not take over and data is missing from Main Met Logger, Snow Temp Logger, and Soil Logger from day 340 to day 351 2001. - 2003 July 2 (DOY 183): Ventilated HMP_24m sensor removed day 183 2003 - 2004 Mar 31 (DOY 91) to August 17 (DOY 230): Soil HeatFlux NE pit #2 10cm appears unresponsive in comparison with other sensors at the same depth from day 91, 2004 to day 230, 2004. - 2004 June 6 (DOY 158) to Sep 19 (DOY 263) (Soil HeatFlux E pit 10cm noisy from day 158 to day 263, 2004. - 2005 August 17 (DOY 229) to August 30 (DOY 242) and September 12 (DOY 255) to December 14 (DOY 348): Diffuse and Total PAR sensors at 25 m reporting low or zero values from 229-242 and from 255 to 348 2005 - 2005 October 6 (DOY 279) to 2006 March 8 (DOY 67): Ventilated PRT air temperature at 24m reporting consistently higher temperatures than other corresponding sensors (~ 6 degrees Celsius). Higher temperatures are likely due to a bad ventilating fan from day 279 2005 to day 67 2006 - 2005 November 28: CR23X datalogger added to reduce dependency on CR7 main met logger. Variables moved from the CR7 to the 23X logger are: Net_RadD_AbvCnpy_20m,Down_PAR_AbvCnpy_25m,Up_PAR_AbvCnpy_20m,DownPAR_BlwCnpy_1m_#2,Air_Temp_Cnpy_6m,Rel_Hum_Cnpy_6m,TcAir_Temp_AbvCnpy_25m,Temp_StrucFunc_AbvCnpy_25m. A small data gap exists for some of these variables on 28 November between 19:30 and 21:00 GMT while sensors were rewired. - 2009 Aug 22 (day 234) 0600 UTC to Sep 3 (day 246) 2100 UTC: all data logged on the main meteorological logger (CR7) was missing. Variables affected included: all Longwave and Shortwave Radiation, 4-way Net Radiation, Downwelling PAR at 1m (#1), BF3 Total and Diffuse PAR, HMP Air Temp and RH at 25m and 1m, Ventilated Thermocouple at 24m, NE Soil Temp profile, NW Tree Boles at 2m, 3m and 5m, Wind Speed at Belfort 3000 rain gauge, Wind Speed and Direction at 26m and Water Table Depth. - 2009 Aug 24 0800 UTC to Aug 28: all data logged on the soil logger was missing, this affected all VWC data. - 2009: Ventilated PRT Air Temp at 24m was excluded for all of 2009, as the sensor was bad and/or showed poor agreement with other sensors for the entire period. - 2009 May to June: Tipping bucket rainfall was excluded because the sensor was broken and was not recording known precipitation events during this period. - 2009 Jan 1 at 0030 UTC to Jan 16 at 2230 UTC: Wind speed at the Belfort 3000 rain gauge was excluded because the sensor was stalled during this period. - 2009 Mar 24 (Day 83) at 1830 UTC to 15 Apr (day 105) at 1630 UTC: Ventilated Thermocouple Air Temp at 24m was excluded because the ventilator fan was stalled and the temperature was showing poor agreement with other sensors. --------------------------------------------------------------------------- 11 . Software 11.1 Software Description [Describe all software that was used to process the data.] Various kinds of plotting and housekeeping software were used to view and process BERMS meteorology data. Some software was commercially available (like PC208 from Campbell Scientific), while others were programs written in C++ and Matlab. Here is a list of current programs and their functions: *** To be written *** 11.2 Software Access [Describe any software that may be available for use by someone who may want to perform further processing of the data. Also describe where a user can get it -- commercial source, Web site, FTP archive, e-mail to author, etc.] Please feel free to contact the following people to discuss software availability and usage: Steve Enns Phone: (306) 975-5683 Email: Steve.Enns@ec.gc.ca Alan Barr Phone: (306) 975-4324 Email: Alan.Barr@ec.gc.ca --------------------------------------------------------------------------- 12 . References 12.1 Platform/Sensor/Instrument/Data Processing Documentation [List any published documentation relevant to the data collected, such as manufacturer's instruction manuals, government technical manuals, user's guides, etc.] Belfort Instrument Company. 1986. Instruction Manual Catalog Number 5-780 Series Universal Recording Rain Gage, Instruction Manual number 8777. Campbell Scientific. 1983. Model 207 Temperature and Relative Humidity Probe Instruction Manual. Campbell Scientific. 1990. SBP270 Barometric Pressure Sensor Instruction Manual. Campbell Scientific Canada Corp. Campbell Scientific. 1992. UDG01 Ultrasonic Depth Gauge Operator’s Manual. Campbell Scientific Canada Corp. Campbell Scientific. 1992. Model HMP35CF Temperature and Relative Humidity Probe Instruction Manual. Campbell Scientific. 1993. CR7 Measurement and Control System Instruction Manual. Campbell Scientific Inc. Campbell Scientific. 1996. CS615 Water Content Reflectometer Instruction Manual. Campbell Scientific Inc. Campbell Scientific. 1996. CSI Model CS700-L Rain Gauge Instruction Manual. Campbell Scientific Inc. Campbell Scientific. May, 1998. Model HMP45C Temperature and Relative Humidity Probe Instruction Manual. Campbell Scientific Corp. Campbell Scientific. Sept, 1998. SR50 Sonic Ranging Sensor Operator's Manual. Campbell Scientific Corp. Carter-Scott Design. No Date. Middleton CN3 Heat Flux Plate Application Note. Carter-Scott Design. May 1995. Instruction Manual for Middleton CN1-R Net Pyrradiometer Edition: CN1R-v1.1. Druck Incorporated. No Date. PTX Depth Pressure Transmitter, Installation and Application Notes. EPLAB, The Eppley Laboratory, Inc. No Date. Instruction Sheet for the Eppley Precision Radiometer (Model PIR). ESI Environmental Sensors Inc. No Date. Moisture Point Instruction Manual. Gray, D.M. 1981. Handbook of Snow. Toronto, ON: Pergamon. Kipp and Zonen. No Date. Instruction Manual Pyranometer CM 11/14. LI-COR. 1991. LI-COR Radiation Sensors Instruction Manual. NASA. 1994. BOREAS Experimental Plan, Version 3. Phillips, D. 1990. The climates of Canada. Ottawa: Canadian Government Publishing Centre. R.M. Young Company. 1980. Instructions, Gill Microvane 3 Cup Anemometer. R.M. Young Co. R.M. Young Company. 1990. Wind Monitor High Resolution Wind Sensor information sheet. R.M. Young Co. Environment Canada. 1993. Canadian Climate Normals 1961-90. Wheaton, E. 1998. But It’s a Dry Cold!. Calgary: Fifth House Ltd 12.2 Journal Articles and Study Reports [List technical reports and scientific publications that concern the methods, instruments, or data described in this document. Publications by the Principal Investigator or investigating group that would help a reader understand or analyze the data are particularly important.] Arain, M.A., T.A. Black, A.G. Barr, P.G. Jarvis, J.M. Massheder, D.L. Verseghy, and Z Nesic. 2002. Effects of seasonal and interannual climate variability on net ecosystem productivity of boreal deciduous and conifer forests. Can. J. For. Res. 32: 878-891. [Abstract] Blanken, P.D., T.A. Black, H. H. Neumann, G. den Hartog, P. C. Yang, Z. Nesic and X. Lee. 2001. The seasonal water and energy exchange above and wthin a boreal aspen forest. Journal of Hydrology. 245(1-4): 118-136. [Abstract] Barr, Alan G., G. van der Kamp, R. Schmidt and T.A. Black. 2000. Monitoring the moisture balance of a boreal aspen forest using a deep groundwater piezometer, Agric. For. Meteorol. 102:13-24. [Abstract] Black, T.A., W.J. Chen, A.G. Barr, Z. Chen, M.A. Arain, Z. Nesic, E.H. Hogg, H.H. Neumann and P.C. Yang. 2000. Increased carbon sequestration by a boreal deciduous forest in years with a warm spring. Geophys. Res. Letters. 29(9): 1471-1274. Chen, W., Black, T.A., Yang, P., Barr, A.G., Neumann, H.H., Nesic, Z., Novak, M.D., Eley, J., Ketler, R., and Cuenca, C. 1999. Effects of Climatic Variability on the Annual Carbon Sequestration by a Boreal Aspen Forest. Global Change Biology, 5(1): 41-53. [Abstract] --------------------------------------------------------------------------- 13. Glossary of Terms and Acronyms [Define discipline-related jargon and the wealth of scientific notations/symbols that may be used in the text, as well all "local" acronyms. Items from the following list may be included. BERMS - Boreal Ecosystem Research and Monitoring Sites BOREAS - BOReal Ecosystem-Atmosphere Study PANP - Prince Albert National Park] MSC - Meteorological Service of Canada (a branch of Environment Canada) BERMS - Boreal Ecosystem Research and Monitoring Sites BOREAS - BOReal Ecosystem-Atmosphere Study BORIS - BOREAS Information System NHRC - National Hydrology Research Centre NWRI - National Water Research Institute SRC - Saskatchewan Research Council UTC - Universal Coordinated Time --------------------------------------------------------------------------- 14 . Document Information 14.1 Document Revision Date [Use yyyy-mm-dd-mmm format] 2006-03-20 2005-06-12 2004-03-22 14.2 Document Author Charmaine Hrynkiw 14.3 Keywords [Include a list of appropriate key words to assist in searching for information.] Meteorology, climate, black spruce, southern boreal forest