«A Thesis Presented to the Faculty of the Graduate School of Cornell University In Partial Fulfillment of the Requirements for the Degree of Master of ...»
Sampling kits were assembled in the lab ahead of the sampling campaign. New needles were included and installed on the syringe in the field. Chamber covers were laid out next to each collar just before the chamber sampling. Before each campaign, the rubber seals on the chamber collars were checked and vegetation moved away from the collar as necessary to allow for proper chamber cover placement.
2. Sample Collection Vial injection: Chamber observations occurred simultaneously at all three transects, with one sampler running a rotating sampling schedule around the transect. For each chamber, the cover was placed over the collar at t=0:00 and the first sample taken from the chamber at t = 0:05.
Subsequent samples were taken at t=20:00, 40:00 and 60:00. Air samples were taken from the chamber by inserting an empty syringe and needle assembly through the rubber septum in the chamber cover. 20ml of air was drawn into the syringe at the appropriate time. The syringe needle was left in the septum for an additional 2 to 3 seconds after the plunger was extended, long enough to allow the pressure to equilibrate between the chamber and the syringe, filling the syringe. The syringe and needle assembly was then withdrawn from the chamber, and the plunger was set to 15 ml. Immediately the needle was inserted into the appropriate vial through
down into the vial. In cases where the plunger was pulled less than 5 ml, it was assumed that the vial had lost its vacuum, and a second withdrawal from the chamber and injection into a spare vial was made, with a note of the adjusted time. After it was verified that the plunger had been pulled at least 5 ml, the full remaining volume in the syringe was pushed into the vial by completely depressing the plunger and the syringe needle was withdrawn from the vials septum.
Upon completing the sampling schedule, we recorded any notes or adjustments.
Soil moisture and temperature: Immediately after the chamber sampling was completed, we observed and recorded soil moisture and temperature at each sampling location. We took three moisture and temperature measurements at each subplot using thermometers and the TDR unit.
The TDR and thermometers were generally placed between the two chamber bases, or to the outside of the pair with care being taken to avoid trampled or disrupted areas.
3. Sample Storage Calibration samples: Immediately after sampling and environmental observations were completed, the calibration vials were injected with the prepared mixes in the same manner as the field samples, labeled and stored in a zip-loc bag.
Storage: All field samples and calibration samples were stored in zip-loc bags within a cardboard box on a shelf in the lab at room temperature until analysis.
Analysis: Just prior to analysis with the gas chromatograph (GC), all field samples and check samples were vented to room pressure by inserting a bare syringe needle through the septum for about 1 or 2 seconds. In most cases, the excess pressure could be heard escaping the vial with a “pshh” sound, indicating that there was still some overpressure in the vial despite the storage time. Each batch of GC analysis was accompanied by calibration samples at three different concentrations.
slopes. Intent was to capture different freezing patterns. Locations were chosen based on apparent wetness or dryness. Low positions were chosen to be near (approximately 2 feet from) standing water in ditches. High positions were chosen based on what appeared to be the driest areas. Soil was mostly frozen. Control soil was almost always frozen, except appeared to be muddy/less frozen very close to standing water. Lower areas in control tended to have uneven frozen ice/mud under the grass, but still solid for the most part. Control high areas tended to be frozen solid as well, but less muddy, all control locations chosen were covered with a good degree of bent-over dead grass. In RC plots, soil seemed less frozen, even thawed in some locations, but also frozen in others. Some locations in RC have a decent layer of dead grass/mulch left from fall mowing while others are much more exposed. Collars were installed by cutting a circle with scissors and removing all easily removed dead grass with as little disturbance as possible and setting dead/dry grass aside. A circular groove was then cut with the saw, this was very difficult and took about 15 minutes per collar. Loose soil frozen and thawed was set aside in a plastic dish, the collar was set into groove and removed soil pressed around outside of collar/soil junction. Disturbed soil inside collar was gently smoothed around the inner edge but not compacted. Dried grass that had been cut was replaced inside the collar. Grass from around the periphery was re-arranged to cover the exposed and compacted area around the outside of the collar so that the layer of insulation was more or less like the surrounding area and no soil was left exposed due to the installation. In a few cases, nearby grass was gathered and placed around the outside of the collar to approximate the surface cover of the immediate area.
4/8/2013: Completed final run yesterday, 12 in total. There was a good sequence of freezing and thawing during sampling period. Large heterogeneity was observed with respect to ice/soil temps
Outside of the collars, I observed large variations in topography, vegetation and soil within the controls, not as much within the treatment, but still somewhat. I conclude that larger chambers would reduce this shielding effect and overcome variation of the surface, and reduce impact of disturbance around collar seat. Also, lower/shorter chambers would be appropriate for this study because there is no standing grass in either control or treatment.
Worked to improve the GC method with success. Essentially, lowering oven temp and extending inlet purge time. New run is longer, but with ~10x sensitivity. Final Method: NOME_20.M Biomass was collected in wet conditions and stored in sealed Ziploc bags in lab till 4/24/2013. I measured total wet mass for each subplot, and pulled a subsample of around 15 g into an aluminum pan. I dried all subplots at once overnight. I weighed dry mass immediately.
5/10/2013: last few days began processing and looking at FT data, first 5 runs. Concentrations look good, and less noisy than older transect runs. Did some preliminary looking at slopes, mostly near zero. Entered soil moisture and temp data into csv files. Now preparing to assemble model and will add remaining flux data when analyzed.
5/31/2013: Looked at recent data and saw a big flux with probable multi-day pattern. Processing raw data by integrating with NOME_20 method on laptop.
41 All original data is provided in comma separated value (csv) format. Some of the data can be used with the provided R scripts to repeat the procedures described in the paper. Brief descriptions of the data heading, and relevance to R scripts is given before each data set.
This data provides the results of the gas chromatography analysis for all field samples and calibrations. It can be processed with script 1 to calculate the chamber fluxes.
“samplecomment”: Indicates additional information about the sample. Check samples • are indicated by “CHK”, calibrations are indicated by “CAL”.