Application of Continuous 222Rn Monitor with Dual Loop System in a Small Lake by Masahiko Ono1, Takahiro Tokunaga2, Jun Shimada2, and Kimpei Ichiyanagi2 Abstract To estimate the spatial distribution of groundwater discharge from the bottom of a small lake of Kumamoto in Japan, we applied continuous radon measurements with a dual loop system and verified the results obtained using the radon method by visual diving surveys. Time-shifting correction in the dual-loop system is reasonable to obtain the true radon activity in water. Distributions of radon activity and water temperature in the study area reveal the effects on groundwater discharge and mixing situation of lake water. The estimated discharge zone ascertained using the radon method agrees with the groundwater discharge distribution observed through diving surveys. Although the data resolution of the radon method is much greater than the width of observed discharge zones, the general distribution of groundwater discharge is recognizable. The dual loop system of radon measurement is useful for smaller areas. Introduction Groundwater has much greater contents of dissolved components than surface water does. It is important to understand that phenomenon and its impact on surface water bodies. Submarine groundwater discharge (SGD) in coastal areas has been recognized as an important pathway from land to ocean (Moore 1996; Burnett et al. 2001a; Taniguchi et al. 2002) and has been widely stud- ied at coastal zones throughout the world (Taniguchi et al. 2002). Groundwater also has a potentially huge impact on lakes. Some studies related to groundwater discharge in saline lakes are reviewed in Zekster (1996). Although groundwater discharge is slight in those saline lakes, it has a strong effect, comparable to that of surface water, on the lake water quality (Zekster 1996). Moreover, in Lake Biwa (fresh water lake), discharging groundwater 1Corresponding author: AIST Tsukuba Central 7, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8567, Japan; +81-29-861-2664;
[email protected] 2Kumamoto University, 2-39-1 Kurokami, Kumamoto 860- 8555, Japan. Received January 2012, accepted October 2012. © 2012, The Author(s) Groundwater©2012, National Ground Water Association. doi: 10.1111/gwat.12002 from the lake bottom has much higher NO3− contents than river water, as observed using an automated seepage meter (Taniguchi and Fukuo 1993; Taniguchi and Tase 1999). Thus, groundwater discharge has significant role in coastal zone of ocean and lake. To evaluate the groundwater discharge in a surface water body, naturally occurring radon has been used as a tracer. Radon (222Rn), a radioactive nuclide, is one nuclide of the 238U decay series. 222Rn, a daughter nuclide of 226Ra, has a half-life of 3.8 d and produces 218Po by alpha decay. Radon is a well-conserved gas that typically shows 2 to 3 orders higher concentration in groundwa- ter than in surface water. As such, it is a good natural tracer for groundwater discharge studies (Burnett et al. 2001b; Burnett and Dulaiova 2003). Continuous radon measurement using a portable radon-in-air monitor was first introduced by Burnett et al. (2001b). More improved method called multi-detector system has been applied for estimation of spatial distribution of groundwater discharge (Burnett et al. 2010; Peterson et al. 2010; Stieglitz et al. 2010). Most of these studies have been conducted on a regional scale and the continuous radon measurement sys- tem has been optimized to that scale. Our focused site is a unique small lake in which groundwater supplies almost all lake water and it is 706 Vol. 51, No. 5–Groundwater–September-October 2013 (pages 706–713) NGWA.org suggested that groundwater discharges from the bottom of lake. To reveal the spatial distribution of groundwater discharge in the lake, we should optimize the continuous radon measurement system for local scale. The aim of this study is to elucidate the spatial distribution of groundwater discharge from the bottom in a small lake by applying continuous radon measurement. Study Area The study area, located in Kumamoto, southwest- ern Japan (Figure 1), had annual rainfall of about 2000 mm/year and average air temperature of 16.9◦C dur- ing 1981 to 2010. Kumamoto is a famous city in relation to groundwater in Japan because its drinking water is almost entirely supplied from groundwater. Many natural springs exist around the city. Kumamoto residents have acted to protect the environment around the springs. The study site, Lake Ezu, is one of “selected 100 spots of clearly preserved water of Heisei Era” and many springs exist around the lake. It has a surface area of 0.5 km2 with water depth of absent around Lake Ezu (Aramaki 2002). Therefore, groundwater of both shallow and deep aquifers discharges into the study area. An earlier study using radon mass balance revealed that deep groundwater is the dominant component of discharged groundwater in the study area (Ono et al. 2011). Results also suggest that groundwater discharge from the bottom of lake exists in the study area and it is unknown (Imatsuji et al. 2008). Method Laboratory Experiment As described previously, continuous radon measure- ment was introduced by Burnett et al. (2001b). A portable radon-in-air monitor (RAD7; Durridge Co. Inc., Biller- ica, Massachusetts), equilibrium chamber, and desiccant are used for continuous radon measurement. An improved method for the observation of spatial distribution of radon activity in water is the multi-detector system, consisting of three RAD7s with parallel connecting tubes (Dulaiova et al. 2005). The multi-detector system has been applied for a study of groundwater discharge in coastal zone and stream (Burnett et al. 2010; Peterson et al. 2010; Stieglitz et al. 2010). Another advanced method from a single- detector system is a dual loop system in Dimova et al. (2009). This method is arranged for measuring thoron (220Rn: half-life 55.6 s) in water while maintaining the effective measurement of radon in water. Two external air pumps are connected to a tubing set of RAD7 in the dual- loop system. Air flow inside the tube is regulated with 3 L/min for the RAD7 chamber and 7.5 L/min for the equilibrium chamber (RAD AQUA; Durridge Co. Inc.), providing good efficiency. Lake Ezu is a small lake (0.5 km2), therefore, we should use a measurement system with a quick response time against the change of radon activity to obtain sufficient spatial resolution. Thoron measurement by dual loop system was expected as ideal method for our study purpose. However, the result of our preliminary survey which measured thoron just above the submergible spring in the lake by using dual loop system is not reasonable for actual distribution of spring. Also we found that change of radon activity would be useful for our case. To confirm the response of the monitoring system, we conducted an experiment with a dual loop system in the laboratory using tap water (high radon) and air (low radon). Tap water was supplied from groundwater at Kumamoto University and water temperature is about 20◦C. The radon activity in the tap water was about 210 dpm/L. Construction of the dual loop system was based on Dimova et al. (2009). Two external air pumps and needle valves were attached to tube of RAD7 unit and equilibrium chamber unit, respectively. Air flow rate is 3 L/min for the RAD7 unit and 7.5 L/min for the equilibrium chamber. Tap water flows into an equilibrium chamber with 2 to 4 min intervals of switching between tap water and air. Water flow rate of 5 L/min was used for our experiment. Table 1 Average Daily Temperatures and Groundwater Head Elevations on the Days of Radon Surveys Groundwater Head Elevation (m) Day Average Daily Temperature (◦C) Kengun A (Shallow) Kengun B (Deep) February 18–19, 2010 5.5 6.37 6.48 June 8, 2010 23.2 6.72 6.58 July 9, 2010 25.3 7.07 6.83 Locations of Kengun Boreholes are depicted in Fig. 2. Field Survey Radon surveys were conducted three times in 2010. Metrological and hydrological conditions on these days are presented in Table 1. The survey was administered in winter with a low groundwater level (February 2010) and summer with high groundwater level (June and July 2010). A survey of Lake Kami-Ezu and Lake Naka-Ezu was conducted on February 18, 2010, and in Lake Shimo- Ezu on February 19. We set two RAD7 devices on a boat. One measured radon in lake water with the dual loop system and the other measured radon in air. Lake water was pumped continuously using a submersible pump into an equilib- rium chamber in 5 L/min flow rate. To pump bottom water into the chamber, the submersible pump location was arranged above the lake bottom by hand regulation each time depending on the measured water depth. The temperature in the equilibrium chamber was measured using a HOBO temperature probe. Water temperature of lake water was recorded using a CTD sensor attached to the submersible pump. For the survey in summer sea- son, some radon grab samples were collected at the study area using another submersible pump and were analyzed for radon activity in collected water using RAD H2O (Durridge Co. Inc.). Radon of grab sample was used for cross-check with continuous measurement. All radon surveys were started at Lake Kami-Ezu after waiting about 30 min for the equilibrium situation. Furthermore, it passed through Lake Naka-Ezu and the west part of Lake Shimo-Ezu. After getting through downstream of Lake Shimo-Ezu (Akitsu Bridge), turned around and passed to the east part of Lake Shimo-Ezu (Figure 2). The boat was paddled by hand to ensure a slow speed for this study. The boat trajectory was recorded using a handheld GPS, showing approximately 5 to 20 m/min speed from recorded data. Diving Survey A diving survey was conducted in December 2010, and in February and August 2011 in the study area (Figure 2). The scuba diving investigations were done in teams of two or three people, with visual observation of 708 M. Ono et al. Groundwater 51, no. 5: 706–713 NGWA.org Figure 2. Trajectory maps showing radon survey and the diving survey. direct groundwater discharges from the lake bottom. The trajectory of the survey and the observed discharge point were recorded using a handheld GPS. Some photographs were taken at discharge points underwater. Results and Discussion Response Time of the Dual Loop System for Radon Activity Change Results of laboratory experiments undertaken to elucidate the response time for change of radon activity in the dual loop system are portrayed in Figure 3. Under changing radon activity from low to high conditions (Figure 3a), it shows a quick response time in both 2 and 4 min intervals. Stieglitz et al. (2010) showed that true radon activity can be achieved by time-shifting correction in the multi-detector system. It is better for understating the presence of radon in water at a precise location. The true radon activity can also be obtained from a dual-loop system by 2 min time-shifting from raw data. For the change of radon activity from high to low (Figure 3b), the response is not clear against change of radon activities. After measuring high radon to low radon, the radon continuous monitoring method shows a “tailing effect,” which takes a long time to recover the radon activity background. The recovery time was reported as 85 min in the single-detector system (Dulaiova et al. 2005). Our result also shows that it seems to take over 30 min to be background level of radon activity. Our laboratory experiment resulted that the increase of radon activity observed from continuous radon Figure 3. Result of laboratory experiment: condition of radon activity change: (a) low to high and (b) high to low. Tap water was running on a 2- or 4-min interval. Measurements of air are shown as blank areas. measurement means an addition of new radon. The dual loop system with measuring in 2-min interval is difficult to know an equilibrium radon activity. However, its quick response of radon increase is sufficient to reveal the dis- tribution of radon activity in lake water and to understand the spatial distribution of groundwater discharge in our study area. NGWA.org M. Ono et al. Groundwater 51, no. 5: 706–713 709 (a) (b) Figure 4. Distribution map of radon activity (a) and water temperature (b) for February 18–19, June 8, and July 9, 2010. Table 2 Average Radon Activity and Water Temperature of Four Spring Waters Around the Study Area Period Average Radon Activity of Spring Waters (dpm/L) Average Water Temperature of Spring Waters (◦C) February, 2010 1472 18.5 June, 2010 1295 19.6 July, 2010 1288 19.2 Radon Activity and Water Temperature in Lake Ezu Distribution maps of radon activity in the lake water are presented in Figure 4a. Radon activity is corrected by 2 min time-shifting from raw data. Radon activity was 200 to 1200 dpm/L. It is high in Lake Kami-Ezu and Lake Naka-Ezu, although it is low in Lake Shimo-Ezu, with no marked seasonal variations between the low groundwater season (February 2010) and high groundwater season (June and July 2010). The averaged radon activity of four spring waters distributed around the study area are 1472 dpm/L on February, 1295 dpm/L on June, and 1288 dpm/L on July (Table 2; Ono et al. 2011). This fact suggests that lake water is strongly affected by groundwater discharge in the Lake Kami-Ezu and Lake Naka-Ezu. Water temperature of the lake water is depicted in Figure 4b. It is 17.0 to 19.0◦C in Lake Kami-Ezu and Lake Naka-Ezu, although it is about 14.0 to 16.0◦C in the Lake Shimo-Ezu on February 2010. In the survey of summer (June and July), water temperature in Lake Kami-Ezu is lower than in Lake Shimo-Ezu. The averaged temperature of the spring water is 18.5◦C in winter and 19.4◦C in summer (Table 2). Therefore, the spatial variation of water temperature and its seasonal fluctuation are caused by and affect the groundwater discharge in Lake Kami-Ezu, with much less influence in Lake Shimo-Ezu. Profiles of Radon Activity, Water Temperature, and Water Depth Profiles of radon activity, water temperature, and depth are portrayed in Figure 5. The horizontal axis shows local time in Japan. The start time of each profile is at a location that is upstream of Lake Kami-Ezu (Figure 2). Radon activity was 800 to 1200 dpm/L in Lake Kami-Ezu. The increasing tendency of radon activity can be recognized (grey shaded areas in Figure 5). Water temperature also changes with increasing radon activity, suggesting that the groundwater discharge from the lake bottom is distributed in Lake Kami-Ezu. Grab samples show similar radon activity at the beginning of the survey. However, because of the tailing effect, radon activity of the grab sample at Saitou Bridge differs from that of the radon survey. A decreasing tendency of radon activity is recognized in the western part of Lake Shimo-Ezu (grey shaded area in Figure 5). In this area, radon activity decreases steeply with increasing water temperature and the changes of 710 M. Ono et al. Groundwater 51, no. 5: 706–713 NGWA.org (a) (b) (c) Figure 5. Profiles of radon activity, water temperature, and water depth in Lake Ezu for (a) February 18 to 19, (b) June 9, and (c) July 8, 2010. Figure 6. Distribution map of the flow velocity at the depth of 1.7 below water surface during summer season in 1998 at Lake Shimo-Ezu (Ohomoto et al. 2001). water depth in June and July, 2010. This tendency is also shown in the result of survey on February 2010, which is no significant tailing effect caused by moving from Lake Naka-Ezu (high radon) to Lake Shimo-Ezu (low radon). The distribution of flow velocity in Lake Shimo-Ezu is presented in Figure 6. It was measured using an Acoustic Doppler Current Profiler at depth of 1.7 m below water surface (Ohomoto et al. 2001). The distribution map of flow velocity clearly illustrates high current flow at western part of Lake Shimo-Ezu, and low current flow in the eastern part, which suggests that lake water in the eastern part has longer residence time than that in the western part. The result must be a much longer time for radioactive decay and atmospheric evasion of radon. Results show that radon activity in lake water of the eastern part is much lower than that of western part. These two different flow patterns might be caused by the mixing of high-radon and low-radon water in the western part of Lake Shimo-Ezu. Longer residence time also affects water temperature in the lake. Water temperature at the lake bottom will be higher in summer and lower in winter than water surface by atmospheric temperature and solar radiation. Therefore, the seasonal fluctuation of lake water at the eastern part is much larger than at the western part. It is higher in summer and lower in winter. Estimation of the Groundwater Discharge Zone The estimated distribution of groundwater discharge zone is shown in Figure 7. Here, we calculated the variation of radon activity against that 2 min prior. If the variation is greater than measurement error, then the zone between a point and previous point is indicated as a groundwater discharge zone with 10 to 40 m spatial resolution. The estimated discharge zones were 11 in February and June, and 6 in July. Results of the diving survey are depicted in Figure 8. A total of 10 zones of direct groundwater discharge from NGWA.org M. Ono et al. Groundwater 51, no. 5: 706–713 711 Figure 7. Estimated spatial distribution of a groundwater discharge zone based on the radon survey and the variation of radon activity. Figure 8. Distribution map of detected groundwater discharge and photograph taken in the diving survey; filled circles denote observed discharge points and the white broken line encloses a discharge point. the lake bottom were found in Lake Kami-Ezu. Further- more, no direct groundwater discharge point exists in Lake Naka-Ezu and Lake Shimo-Ezu. These discharge zones are discharge as estimated by the radon survey can be recog- nized in the study area. The distribution agrees with the results of the diving survey, especially in June 2010 (high- est spatial resolution). This fact strongly suggests that the continuous radon measurement with a dual-loop system is useful even at small study sites. Conclusions Our study is summarized as follows. • The dual loop system is useful for quick response interval of radon activity change. Time shifting of one cycle is reasonable for showing the true radon activity in the water. In conditions of change of high to low radon, a tailing effect must occur. • Distribution maps of radon activity and water temper- ature showed that lake water is strongly affected by groundwater discharge in Lake Kami-Ezu and Lake Naka-Ezu, and by mixing at Lake Shimo-Ezu. The estimated discharge zone is widely distributed in Lake Kami-Ezu in total ten zones. The results of diving sur- veys also support radon distribution. • Data resolution of radon surveys is much greater than the width of observed discharge zones. However, a general distribution of groundwater discharge can be recognized in the study area. This fact strongly suggests that continuous radon measurements using a dual loop system is applicable for small area. Chanyotha et al. (2011) succeeded to find out thoron peak relating to groundwater inflow in Chao Phraya River. We believe that thoron must be effective tracer to under- stand the spatial distribution of groundwater discharge precisely in Lake Ezu. Additional experiment and survey are needed to apply thoron method in the study area. Acknowledgments The authors acknowledge Prof. Ichikawa T. of Tokai University for the introduction of Lake Ezu and his suggestion to our study, and appreciate Mr. Honda K. and the staffs of Kumamoto City Zoological and Botanical Gardens for support of our radon survey. Mr. Yamaura T. and the diving club “Dive Bears” of Kumamoto University greatly helped our diving survey. Waterworks and Sewerage Bureau, Kumamoto City provided us long- term data related to the groundwater level. We appreciate editor and reviewers for careful reading our manuscript and for giving useful comments. References Aramaki, S. 2002. Studies on the groundwater conservation of KITAAMGI spring water in KASHIMA. Report of research project, Grant-in-Aid for Scientific Research, 12680582. Burnett, W.C., M. Taniguchi, and J. Oberdorfer. 2001a. Mea- surement and significance of the direct discharge of groundwater into the coastal zone. Journal of Sea Research 46: 109–116. Burnett, W.C., G. Kim, and D. Lane-Smith. 2001b. A continu- ous monitor for assessment of 222Rn in the coastal ocean. Journal of Radioanalytical and Nuclear Chemistry 249, no. 1: 167–172. Burnett, W.C., and H. Dulaiova. 2003. Estimating the dynam- ics of groundwater input into the coastal zone via contin- uous radon-222 measurements. Journal of Environmental Radioactivity 69: 21–35. Burnett, W.C., R.N. Peterson, I.R. Santos, and R.W. Hicks. 2010. Use of automated radon measurements for rapid assessment of groundwater flow into Florida streams. Journal of Hydrology 380: 298–304. Chanyotha, S., M. Taniguchi, and W.C. Burnett. 2011. Detecting groundwater inputs into Bangkok canals via radon and thoron measurements. In Groundwater and Subsurface Environments: Human Impacts in Asian Coastal Cities, ed. M. Taniguchi, 145–158. New York: Springer. Dimova, N., W.C. Burnett, and D. Lane-Smith. 2009 Improved automated analysis of radon (222Rn) and Thoron (220Rn) in natural waters. Environmental Science and Technology 43: 8599–8603. Dulaiova, H., R. Peterson, W.C. Burnett, and D. Lane-Smith. 2005. A multi-detector continuous monitor for assessment of 222Rn in the coastal ocean. Journal of Radioanalytical and Nuclear Chemistry 263, no. 2: 361–365. Imatsuji, G., T. Ichikawa, and S. Aramaki. 2008. On the Spring Rate Change and those mechanisms in the EZU Lake and SUIZENJI KUMAMOTO. Bulletin School of Industrial Engineering Tokai University 1: 46–52. Kumamoto Prefecture and Kumamoto City. 1995. Written report on a comprehensive survey of groundwater in Kumamoto district, 127. Moore, W.S. 1996. Large groundwater inputs to coastal waters revealed by Ra-226 enrichments. Nature 380: 612–614. Ohomoto, T., K. Yakita, and H. Fukushima. 2001. Horizontal convection and water quality characteristics in Lake Ezu with spring water. Journal of Hydroscience and Hydraulic Engineering 45: 1183–1188. Ono, M., J. Shimada, T. Ichikawa, and T. Tokunaga. 2011. Eval- uation of groundwater discharge in Lake Ezu, Kumamoto, based on radon in water. Japanese Journal of Limnology 72: 193–210. Peterson, R.N., I.R. Santos, and W.C. Burnett. 2010. Evaluating groundwater discharge to tidal rivers based on a Rn-222 time-series approach. Estuarine, Coastal and Shelf Science 86: 165–178. Stieglitz, T.C., P.G. Cook, and W.C. Burnett. 2010. Inferring coastal processes from regional-scale mapping of 222Radon and salinity: examples from the Great Barrier Reef, Australia. Journal of Environmental Radioactivity 101: 544–552. Taniguchi, M., and Y. Fukuo. 1993. Continuous measurements of ground-water seepage using an automatic seepage meter. Ground Water 31, no. 4: 675–679. Taniguchi, M., and N. Tase. 1999. Nutrient discharge by ground- water and rivers into Lake Biwa, Japan. In Proceedings of IUGG 99 Symposium, IAHS Publication No. 257, 67–73. Taniguchi, M., W.C. Burnett, and J.E. Cable. 2002. Investigation of submarine groundwater discharge. Hydrological Pro- cesses 16: 2115–2129. Yoshikura, M. 1986. Nature of Lake Ezu, series of natural environment in Kumamoto, ed. M. Yoshikura. Kumamoto Biological Laboratory, 112. Zekster, I.S. Groundwater discharge into lakes: a review of recent studies with particular regard to large saline lakes in central Asia. International Journal of Salt Lake Research 4: 233–249. NGWA.org M. Ono et al. Groundwater 51, no. 5: 706–713 713