Much of the early climatological measurements on the Andrews Forest were associated with the investigation of the initial small watersheds experiment (Watersheds 1, 2, 3). The Climatic Station on WS 2 was installed in 1956 to measure precipitation and a hygrothermograph was added in 1958. The High-15 station was established to measure precipitation in 1964 in conjunction with small watersheds 6, 7, 8, and has evolved into a secondary meteorological station. A more general set of modeling needs led to the installation of the Primary Meteorological Station in 1972 to characterize the meso-scale environment. Originally, solar radiation, air temperature, dew point temperature, and windspeed were collected. Along with precipitation from the climatic station on WS 2, these were the primary climatic variables needed for the models predicting the rates at which materials accumulate or move through ecosystems (Waring et al., 1978). Significant improvements to the station were made in 1975, 1979, and 1988, as the station evolved from chart recorders to state-of-the-art digital data loggers (Bierlmaier and McKee, 1989). The Vanilla Leaf Met Station was installed in 1987. The primary intent was to provide micro- meteorological data for a study of seedling survival following clearcut and shelterwood logging at high elevation. Ultimately, the shelterwood site was discontinued and the clearcut site has evolved as a primary high elevation meteorological station.
Four Benchmark Meteorological Stations (BMS) and two second-level stations are included in the MS001 database. The Primary Meteorological and Vanilla Leaf Meteorological BMS are retained. Two new BMS are installed. In 1994, the Upper Lookout Meteorological Station was established at high elevation (4200 ft, ENE aspect) on clearcut L708 in the SE Andrews. In 1995 the Central Meteorological BMS was established at a centrally located site on clearcut L351 (3300 ft, WSW aspect) in the east-central Andrews. A GIS analysis of elevation and aspect indicated the average elevation (3170 ft., 966 m) and average aspect (267 degrees) of the Andrews Forest, and the Central Met Station was located to represent these general averages. Modifications are made to the Primary and Vanilla Leaf Stations to standardize measured variables, temporal resolution, methods, and instrumention across all BMS. Sites will be cleared and required openings maintained following standards of the National Weather Service, the LTER network, and where appropriate, the NADP network. Telemetering of all BMS was completed in 1996. Second-level stations (SLS) at the Hi-15 and WS 2 Climatic Station will continue to be maintained for measurement of precipitation, temperature, and other data to maintain continuity of historical records. These sites also follow established procedural standards.
The BMS include meteorological measurement of air and soil temperature, relative humidity, calculated dew point temperature and vapor pressure deficit, wind speed and direction measurement, incoming solar radiation, photosynthetically active radiation (PAR), soil moisture, snow melt, and snow moisture and depth.
Please follow this link to see a general description and history of air temperature measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/air.htm
Temperature data has been measured with a variety of instruments:
Probes were housed in locally designed PVC radiation shields from the 1980s until the mid 2000s. Gill radiation shields replaced the PVC shields at all of the benchmark sites after a significant concurrent comparison was made. Aspirated shields are also now in use. The original 1.5 meter Primary Met Station sensor was housed in a Cotton Region Shelter. Historical methods included Rustrak strip charts with thermister where air temperature is measured by a Yellow Springs Instruments linear thermistor in a standard Cotton Region Shelter and continuously recorded on a separate 30-day Rustrak strip chart scaled from -10 degrees C to 40 degrees C.
Data loggers are typically Campbell Scientific (CR10, CR21X, CR500, or CR23X). Early data loggers used in the 1980s are Interface Instrument models M-2, M-3, and M-4.
Please follow this link to see a general description and history of dew point temperature measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/dew.htm
Dew Point is calculated from air temperature and relative humidity every sampling interval (generally every 15 seconds) and output hourly using Tetens Equation for the relation between temperature and the partial pressure of water vapor (see http://andrewsforest.oregonstate.edu/data/studies/ms01/dewpt_vpd_calculations.pdf). This is calculated for independent pairs of air temperature and relative humidity probes at 150 and 450 cm heights for all benchmark stations (except CS2MET which only has sensors at 150 cm height).
Historical methods at PRIMET used a heated lithium-chloride dew point sensor with a thermistor that recorded continuously on a separate 30-day Rustrak strip chart throughout the 1970s. An Interface Instrument hygrometer with linear thermister was used to measure dew point indirectly with a heated Interface Instrument lithium-chloride hygrometer with a linear thermistor from 1979 until 1988. The PRIMET early dew point sensors and subsequent air temperature and relative humidity sensors used to calculate dew point were located in the 150 cm height cotton shelter until May 2000. The cotton shelter was discontinued for dew point calculation and replaced with use of the 450 cm probe on the PRIMET tower.
Please follow this link to see a general description and history of precipitation measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/ppt.htm
Precipitation records are generally evaluated every 5 minutes and data points are output to capture 5 minute resolution. In actuality, not all of the gauges are designed for accurate recording at this high temporal frequency except during periods of high intensity rain/snowfall. This is generally due to the sensitivity of the magnetostrictive tank gauge and associated tank diameter in measuring fluid level. The "stand-alone" gauges at UPLMET, VANMET and CENMET are designed for greater sensitivity to changes in tank depth and the 5 minute output is good during most precipitation events. The "shelter gauges" at those sites are not as sensitive to a change in fluid depth. The tipping bucket gauge at PRIMET reliably outputs tips on a 5 minute basis, although this record is somewhat artificial at 5 minute resolution, especially in light rain situations where the gage does not tip in every 5 minute segment and 15 minute resolution is probably more realistic. Early chart data at CS2MET and the H15MET have been digitized and are generally accurate at 15 minutes resolution.
The longest continuous precipitation record on the Andrews is at the Climatic Station on Watershed 2 (CS2MET): 1957-10-01 to Present. Data is collected weekly from a standard US Weather Bureau standard 8” rain gauge and an adjacent Universal recording gauge (chart). The weekly charts were originally hand digitized on a daily basis, and adjusted to the standard gauge total. Subsequently, all charts have been electronically digitized, and data can be reliably output at a 15 minute resolution. This high resolution chart data has also been corrected to the standard gage total on a weekly basis. This Universal rain gauge was originally positioned at another location on the Andrews (at FORKS) and that record can be used to extend the daily CS2MET precipitation record back to December, 1951. A ETI Instruments NOAH IV gauge was installed in 2011 and runs concurrently with the Universal gauge and the standard gauge
At the Hi-15 Meteorological Station (H15MET), a standard rain gauge has always been maintained along with an accompanying recording rain gauge. Historically, the charts were hand digitized on a daily basis, and the daily record corrected to the standard gauge periodic total using a prorating technique. Beginning in 1992 a more reliable recording gauge was installed using a pressure transducer to measure fluid depth, and this record is available with no correction to the standard gauge, although the standard gauge is still operational.
At both the Upper Lookout (UPLMET) and Central (CENMET) Stations, two gauges are in operation: a stand-alone gauge with two concentric, encircling wind fences, and a second gauge atop the station shelter. The orifice of the stand alone gauge is heated by a thermistatically-controlled propane heater. This heated gauge with the shielding fences has dramatically improved our precipitation catch, particularly in heavy snow conditions, and has prevented snow bridging over the gauge orifice. The shelter gauge is located on top of our station shelter and heated through ducts from the shelter's propane heating system. This gauge has also improved catch over the more traditionally heated rain gauges, but is protected only with an alter wind shield and catches less than the stand-alone.
At Vanilla Leaf (VANMET) the original gauge in the clearcut was installed in 1987 but the record is discontinued in 1996. This gauge suffered from its openly exposed siting in this high elevation clearcut. Heavy snow and inadequate wind shielding were major problems. A stand-alone type gauge similar in design to those used at UPLMET and CENMET was installed in the adjacent and naturally sheltered meadow above the VANMET clearcut in 1998. This site takes advantage of the naturally-protected meadow and the wind fences are not employed here.
Please follow this link to see a general description and history of relative humidity measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/rel.htm
Generally, relative humidity is output hourly and is measured at heights of both 150 and 450 cm at these benchmark stations using Vaisala sensors with Campbell data loggers. Early records at CS2MET employed a hygrothermograph sensor to record daily max-min relative humidity from within a Cotton Region shelter on a recording chart from 1958 to 1998. Standard maximum and minimum thermometers on a Townsend support and a sling psychrometer were used to make weekly corrections to the hygrothermograph chart readings.
Please follow this link to see a general description and history of snow lysimeter measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/lys.htm
Precipitation and melting snow runoff are measured by a tipping bucket with tips summarized in 5 minute intervals. General installation: The lysmeter collection pan was constructed from 2 x 12s and plywood. The pan was covered with hyplon (white vinyl polymer), and a bathtub drain installed in one corner. The pan was installed on the ground surface and was leveled such that the collected water would flow to the corner with the drain. The water is then directed to a tipping bucket (TB) in the basement of the structure by pvc pipe which is buried to prevent freezing. The final deminsions of the lysimeter pan are 92" x 93" x 12", so a volume of water equal to .01" of ppt is 1.4032 L.
Please follow this link to see a general description and history of snow water equivalence (moisture) and snow depth measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/sno.htm
Snow water equivalence (SWE) and snow depth are measured every 5 minutes. Due to high variability in readings, particularly of snow depth, values may only be posted every hour for some data. Snow moisture is measured with a Park Mechanical pressure pillow and Druck pressure transducer. Snow depth is measured with a Campbell Scientific model SR50 sonic ranging snow depth sensor or Judd communications snow depth sensor. The snow depth sensor takes up to 10 readings, performs error checking and outputs a reading in mm every 5 minutes. Depth sensor measurement range is 0.5 to 10 meters. The snow sensor is mounted on a long pipe extending from the shelter. The snow pillow readings are instantaneous at 5 minute intervals. Campbell Scientific data logger is used. Snow moisture and snow depth are merged in the output files.
Snow moisture and snow depth are measured with snow cores at 5 points across a transect (see entity 20). This station snow course (transect) is near the station and tries to follow a single contour. The average of these 5 points provides a check of the snow pillow and snow depth sensor. Snow depth is also measured at each corner of the station snow pillow using a permanently set graduated stake (pvc pipe). The average depth is converted to snow water equivalence using the density calculated from the station snow course.
Please follow this link to see a general description and history of soil moisture measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/smo.htm
Soil moisture is measured with a water content reflectometer (WCR) beginning in 1998 replacing older gypsum block technology. Values are output on a daily basis at four depths (10, 20, 50, 100 cm) from the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET)
Please follow this link to see a general description and history of soil moisture potential measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/smp.htm
Gypsum blocks were used to record mean daily soil moisture potential at the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET) beginning in 1994. Water content reflectometers were added beginning in 1998 at all sites and all gypsum block measurements were terminated by May 2000. Hourly measurements were made between 1988 and 1990 at PRIMET but were discontinued and restarted in Oct 1994 on a daily basis only. At VANMET, soil moisture potential measurements were summarized by daytime and nighttime hours only from 1987 to 1994 at depths of 10, 20, 30 cm. From 1987 to 1993, mean daily soil moisture potentials were recorded in the rooting zone of 3 individual trees at VANMET.
Please follow this link to see a general description and history of soil temperature measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/soi.htm
Mean hourly soil temperatures are measured at the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET) at four depths (10, 20, 50, 100 cm). Thermocouple wire was used in the earliest measurements and were protected inside of a PVC structure. Most measurements begin in 1994 but measurements were made at PRIMET beginning in 1988 at 20 cm depth only. At VANMET, measurements were summarized by daytime and nighttime hours only from 1987 to 1993 at depths of 10, 20, 30 cm, and soil temperature was recorded in the rooting zone of 3 individual trees.
Please follow this link to see a general description and history of photosynthetically active radiation (PAR) measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/par.htm
Mean Photosynthetically active radiation (PAR) is measured at Central Meteorological Station (CENMET) and output at 15 minute intervals. Maximum solar radiation is based on instantaneous micromoles per second per square meter, and the time of the maximum value is accurate to the minute. It is common for solar radiation sensors to read negative at night (-1 to -7), and negative nighttime readings are changed to 0. The daily values average in these negative values, creating a minor error in the daily values.
Please follow this link to see a general description and history of solar radiation measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/sol.htm
Total incoming solar radiation is recorded at the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET) and output every 15 minutes starting in 1994. This total incoming solar radiation data is also output every day in megaJoule per m2 along with the maximum daily rate of incoming solar radiation in watts per m2 and the time of the maximum accurate to the minute. Solar radiation measurement began in 1972 at PRIMET, but much of the incoming solar radiation data from the 1970's is estimated. Values are output on an hourly basis until 1994 at PRIMET. Data had originally been output as langleys but has been converted to more standard units.
Please follow this link to see a general description and history of water vapor pressure deficit measurements for these stations:
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/vpd.htm
Vapor pressure deficit values are calculated and output hourly from air temperature and relative humidity probes that sample every 15 seconds. Values are output in millibars and are collected at 150 and 450 cm heights at each of the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET). The earliest hourly calculations were started at PRIMET in 1988 from air temperature and relative humidity sensors located in the 150 cm cotton shelter based on sampling every 15 seconds (in millibars). Please see the following for calculation details: http://andrewsforest.oregonstate.edu/data/studies/ms01/dewpt_vpd_calculations.pdf
http://andrewsforest.oregonstate.edu/data/studies/ms01/meta/wnd.htm
Hourly mean wind speed, resultant mean wind vector magnitude and standard deviation, resultant mean daily wind vector direction, and daily wind direction vector components (Wind Rose) are output at each of the four benchmark stations (PRIMET, CENMET, UPLMET, VANMET). The data logger samples every 15 seconds and the sensor is installed a 10m height. It has been discovered that wind speed from this propeller-style wind anemometers are not accurate below 1 meter per second. Much of this data will be tagged as 'Questionable' or possibly removed.
Dewpoint and vapor pressure deficit equations can be found at:
http://andrewsforest.oregonstate.edu/data/studies/ms01/dewpt_vpd_calculations.pdf
A description of the early processing history and instrumentation for the Primary Meteorological Station (PRIMET), 1972-1989, is posted here:
http://andrewsforest.oregonstate.edu/data/studies/ms01/ms001_primet_1972_1989.pdf
