Six years of atmospheric CO&amp;lt;sub&amp;gt;2&amp;lt;/sub&amp;gt; observations at Mt. Fuji recorded with a battery-powered measurement system

Shohei Nomura,Toshinobu Machida,Yukio Terao,Hitoshi Mukai,Yukihiro Nojiri

doi:10.5194/amt-10-667-2017

Abstract

Abstract. We developed a battery-powered carbon dioxide (CO2) measurement system for monitoring at the summit of Mt. Fuji (3776 m a.s.l.), which experiences very low temperatures (below −20 °C) and severe environmental conditions without access to gridded electricity for 10 months (from September to June). Our measurement system used 100 batteries to run the measurement unit during these months. These batteries were charged during the 2-month summer season when gridded electricity was available, using a specially designed automatic battery-charging system. We installed this system in summer 2009 at the Mt. Fuji weather station; observations of atmospheric CO2 concentration were taken through December 2015. Measurements were never interrupted by a lack of battery power except for two cases in which lightning damaged a control board. Thus we obtained CO2 data during about 94 % of the 6-year period. Analytical performances (stability and accuracy) were better than 0.1 ppm, as tested by checking working standards and comparisons with flask sampling.Observational results showed that CO2 mole fractions at Mt. Fuji demonstrated clear seasonal variation. The trend and the variability of the CO2 growth rate observed at Mt. Fuji were very similar to those of the Mauna Loa Observatory (MLO). Seasonally, the concentration at Mt. Fuji was 2–10 ppm lower in summer and 2–12 ppm higher in winter than those at MLO. The lower concentrations at Mt. Fuji in summer are mainly attributed to episodes of air mass transport from Siberia or China, where CO2 is taken up by the terrestrial biosphere. On the other hand, the relatively higher concentrations in winter seem to reflect the high percentage of air masses originating from China or Southeast Asia during this period, which carry increased anthropogenic carbon dioxide. These results show that Mt. Fuji is not very influenced by local sources but rather by the sources and sinks over a very large region.Thus we conclude that, as this system could provide stable measurement data with relatively easy operation for 6 years at Mt. Fuji, it could be a useful monitoring technique for remote background sites elsewhere.

Highlights

In order to quantitatively understand the global carbon cycle, long-term observations of atmospheric CO2 concentrations have been performed at “background sites” where the observed air is less affected by local sinks and sources, such as Mauna Loa in Hawaii and Antarctica (Keeling et al, 1989, 2001; Conway et al, 1994)
We developed a CO2 measurement system that can be operated without gridded electricity, even under the harsh conditions found at the summit of Mt
The calibration using three working standard gases was done repeatedly every day. These working gases were calibrated by our NIES calibration system in the laboratory with a high accuracy below 0.1 ppm, which was confirmed by a World Meteorological Organization (WMO)/International Atomic Energy Agency (IAEA) Round Robin inter-comparison

Summary

Introduction

In order to quantitatively understand the global carbon cycle, long-term observations of atmospheric CO2 concentrations have been performed at “background sites” where the observed air is less affected by local sinks and sources, such as Mauna Loa in Hawaii and Antarctica (Keeling et al, 1989, 2001; Conway et al, 1994). The importance of observations at “regional sites” has been recognized when analyzing regional sinks and sources using an inverse model with data from various ground observation sites, such as Siberia and China (Maksyutov et al, 2003; Saeki et al, 2013; Zhang et al, 2014). Representative sites for the long-term observation of atmospheric CO2 concentrations in Japan are located in the higher- and lower-latitude coastal areas of Ochi-ishi, Minamitori Island, Hateruma Island, and Yonaguni Island (Fig. 1; Watanabe et al, 2000; Mukai et al, 2001)

Methods

Results

Conclusion