Running your Picarro GasScouter and eosAC Off-Grid – Part 3

Introduction

We deployed the GasScouter and eosAC system at Dalhousie University’s Bio-Environmental Engineering Centre (BEEC).  BEEC is a research and demonstration site operated jointly by the Faculty of Agriculture’s Engineering Department and Dalhousie University’s Department of Biological Engineering.

We set up the system on a constructed wetland at the BEEC site (normally used for studying filtering of biosolids) with the purpose of monitoring the CO2 and methane emissions from the wetland in comparison with the emissions from the raised land near the wetland (where the shed and equipment was set up – which we’ll call the upland from now on).

Equipment Setup

Our shed, solar panels and batteries were set up – as we described in Blog Post 2 – on the upland portion of the site.

GasScouter and eosAC flux chamber shed

We installed four eosAC chambers at the site, three in the wetland soil (approximately 30-40 cm from each other) and one on the upland, as seen below. Each chamber had a 10 cm tall soil collar, inserted approximately 7 cm into the soil (3 cm exposed lip) and 15 meters of tubing from the chamber to the shed containing the GasScouter and eosMX chamber multiplexer.

eosAC ChamberseosAC Chambers

Using the eosLink-MX software, the chambers were set to each do 10 minute long measurements in sequence (1,2,3,4) with a 60 second purge time before chamber closure.

Performance and Preliminary Flux Data

We visited the field site approximately 5 days after the initial deployment and found the system was 50% working (just the eosMX-P in this case). After some troubleshooting of the system we had found that the GasScouter suffered from a blown fuse in its internal battery; probably because of a low voltage in the deep-cycle batteries after a few days of cloud, causing a large increase in required charge current. In order to prevent this in future, Darren did a quick calculation and determined that we should put a fast-blow 7 Amp fuse in line with the GasScouter power. This way, if the GasScouter tries to draw too much current, the 7 Amp fuse blows rather than the fuse in the Lithium Ion battery pack (not user replaceable).

After repairing the fuse in our electronics shop, we took the system back to the field for the weekend to gather a bit more data before it had to be pulled for other uses.

During the course of the experiment, we observed a temperature correlated CO2 flux in all of the plots; although the upland chamber (Chamber 1) demonstrated a weaker temperature response likely due to very dry soil.

Time series fluxes of carbon dioxide and chamber temperature during the deployment period

Fluxes of carbon dioxide and chamber temperature during the deployment period. Note the strong positive correlation between temperature and CO2 flux.

Chamber temperature versus flux for the wetland (blue) and upland (green) showing the upland had a reduced temperature response because of dry conditions.

Chamber temperature versus flux for the wetland (blue) and upland (green) showing the upland had a reduced temperature response because of dry conditions.

There was consistent methane uptake from the upland chamber, whereas the wetland chambers showed consistent methane emission with an average rate of 19.85 nmol/m2/s and a standard deviation of 54.7 nmol/m2/s with a positively skewed distribution.

Chamber methane fluxes for the wetland (blue, red, green) and upland (black) sites. Note that the wetland is plotted on the left y-axis and is always emitting whereas the upland is plotted on the right y-axis and has a consistent uptake.

Chamber methane fluxes for the wetland (blue, red, green) and upland (black) sites. Note that the wetland is plotted on the left y-axis and is always emitting whereas the upland is plotted on the right y-axis and has a consistent uptake.

During our second deployment we had a small rainfall event that drove massive increases in methane flux for the wetland chambers (see above), but with almost no change in the methane dynamics for the upland chamber. It’s unclear if this was an event that caused a short-term increase in the production of methane, or simply drove methane out of the wetland soil pores by volume replacement. Some of the flux curves suggest ebullition of methane, which is more consistent with the volume replacement theory.

eosAnalyze-AC screenshots showing a well behaved increase in CO2 concentration (left) and a CH4 time series that is consistent with ebullition (right).

eosAnalyze-AC screenshots showing a well behaved increase in CO2 concentration (left) and a CH4 time series that is consistent with ebullition (right).

That’s all for the data for now. We’ll be presenting the fully worked up data at the ASA, CSSA & SSSA International Annual Meeting in Phoenix from November 6-9. The poster will be in the “General Wetland Soils Poster 1” session from 14:30-16:30 on Wednesday, November 9th.

Conclusion

Our enclosure and solar panel system were able to keep the system dry and powered, although an increase in the safety factor on both the batteries and solar supply is recommended as it’s not guaranteed that the solar for your site will be consistent with the mean values we presented in Blog Post 1.

Luckily we experienced the low-voltage fuse issue ourselves, and now can recommend that users install and additional fuse (fast-blow, 7 Amps) before the GasScouter to protect its internal fuses and electrical components.

The data shows expected trends, including a positive correlation between CO2 flux, temperature and methane release in the wetland. Interestingly we observed very high rates of methane release after a single rain event, which could be caused by release of methane from the soil pore space during percolation of rain water.

Acknowledgements

Thanks to David Burton and Peter Harvard for hosting us at the BEEC site during this study; and to David Kim-Hak and the technical team at Picarro for their technical assistance during the deployment.