In this section, manipulations and adjustments to the measured data are summarized for each variable.
The 1992 air temperature values have been adjusted slightly, based on a four day post-field season intercomparison of the sensors used at Delta and Echo. Using an average difference in the comparison period values of 0.16°C, Delta values were adjusted up 0.08° C, and Echo values down 0.08° C.
At the end of each day the maximum and minimum values were recorded, from the 1440 values measured. The times at which the daily maximum and minimum temperatures occurred were also recorded. Daily mean temperatures were calculated as the average of all 1440 values, rather than by averaging the daily extremes. The daily maximum and minimum temperatures are based on daily extreme hourly mean values for 1 day at Echo in 1991, due to a micrologger program change. In 1992 there were 3 maximum and 6 minimum daily temperatures determined this way. The maximum and minimum temperatures for these days are therefore conservative, and the times of maximum/minimum occurrence are for the extreme hourly means.
The 1992 RH values have been adjusted slightly, based on a four-day post-field season intercomparison of the sensors used at Delta and Echo. Based on an average difference in the values of 1.2 percent, Delta values were adjusted down 0.6 percent, and Echo values up 0.6 percent.
Echo wind speed and direction in 1991 was measured every minute and vectors computed every hour, as discussed in the data acquisition section. There were several time periods however, when the sensors are either known or suspected to have been rimed to such extent that they ceased to function. For six such periods, ranging between 1 and 28 hours, the data are missing due to complete immobilization of one or both sensors.
The 1992 Echo wind data was collected with two execution intervals (20 and 60 seconds), in support of a snowpack ventilation experiment. There were also a couple of problem periods, due to programming errors, a possible bad channel on the micrologger, and riming. A table is available which summarizes the format of wind data collected during the 1992 field season. (TABLE: ECHO wind measurements 1992)
Wind direction was referenced to true north. An attempt was made each year to orient the sensor using its shadow at solar noon (local apparent time), with varying degrees of success. During both seasons, alignment was checked by recording the shadow orientation at a specific date and time. After the field season, solar azimuth was computed based on latitude, date and the observation time. The difference between the predicted and recorded shadow orientation in all but one case ranged from 2.5° to 8.8°, and adjustments were not made. The exceptional case was the Echo sensor in 1991, whose position had been established by reckoning, due to persistent cloud cover. In this case the difference was 21.3°, and an adjustment, or offset, was made to the wind vector direction.
No measurements made at Echo station.
In 1992, an electrical problem was apparently associated with the K-up radiometer wiring harness plug at Echo, and resulted in loss of some data between June 1 and July 7. Generally, the problem occurred during wet weather, and persisted from 1 to 6 hours. In one case, data were lost for a 48 hour period. Bad values of K-up were determined by plotting K-down and K-up at Echo (including bad data), along with both terms at the Delta site, and albedo computed from raw Echo values. Aberrant K-up values resulted in albedo deviations that were easily recognized on the plots, in conjunction with weather observations and Echo station notes. Replacement values were computed by interpolating albedo values and multiplying these by values of K-down.
No measurements made at Echo station.
Computation of L-down followed the method of Wardle and McArthur (1992) using the pyrgeometer signal, adjusting for ambient temperature, and correcting for short-wave interference:
where Vthp and Vg are the pyrgeometer and pyranometer signals, respectively, in mV; R and Rg are the responsivities, in mV W^-1 m^-2, of the pyrgeometer and pyranometer; E is the Stefan-Boltzmann Constant (5.67*10^-8 W m^-2 K^-4); and T is the absolute temperature of the pyrgeometer determined from the case thermistor (° K; see below). Although this equation was developed for pyrgeometers mounted inside NARC ventilated housings (the pyrgeometers at Taconite Inlet were not ventilated, due to logistical considerations), the method is preferable over the battery compensation circuit for unventilated instruments as well (McArthur, personal communication). The responsivity used for each pyrgeometer was the average of that determined in calibration before and after the field season. Comparison computations of L-down using the two different calibration values resulted in differences of only 1.5 percent (n = 60 hourly values), so averaging only altered the results slightly.
Field measurements of pyrgeometer case thermistor resistance were also made, using a DC half bridge (YSI model 44031 precision thermistor). These resistances were used to determine the pyrgeometer case temperature, allowing a correction to be entered into the L-down computation accounting for changes in ambient temperature. The equation used is from Wardle and McArthur (1992):
where T is the required absolute temperature (° K); To is 273.15 ° K + 25 ° K; LGr is Loge(r/10000); and r is the measured resistance, in ohms. At Delta early in the season, a programming error resulted in loss of resistance measurements for a period of 70 hours. To compensate, a regression equation was determined for predicting case temperature from air temperature during similar weather (r^2 = 0.85), and used to estimate missing case temperature values.
Computation of thermal exitance from the surface (L-up) followed a modification of the Stefan-Boltzmann Law, for less than full radiators:
where the emissivity coefficient G is taken as 0.99 for snow (Dozier and Warren, 1982), and 0.95 for tundra (cf. Ohmura, 1980); the Stefan-Boltzmann Constant E is 5.67*10^-8 W m^-2 K^-4, and T is the adjusted surface temperature in ° K. The appropriate emissivity coefficient was determined by the albedo at each measurement time: when albedo > 0.5 the surface was classified as snow covered.
No adjustments were made, except as noted above for the K-up component.
Precipitation was measured only a the Delta station, and manipulation of the data is discussed in the documentation for that station.