Please use this identifier to cite or link to this item: http://hdl.handle.net/1946/13001
Application of soil measurements and remote sensing for monitoring changes in geothermal surface activity in the Reykjanes field, Iceland
The objective of this study was to summarize the available data on the surface activity in the Reykjanes geothermal area for the last eight years and reveal the changes which have taken placethere during this period and also to evaluate the methods which have been used. Annually since 2004, soil measurements werecarried outon soil temperature at 15 cm depth and CO2 emission through soil on a measurement grid with a 25 x 25 m grid spacing except in 2011 when two such datasets were made. Total heat flow, total CO2 flux and uncertainty was calculated for each year’s measurements and the distribution shown on maps. A thermal infrared image, obtained in May 2011 from Reykjanes, snowmelt tracks, done in March 2011 and results from temperature loggers which obtained data during four periods from May 2011 to April 2012 were also used. The soil measurements show that heat flow has increased from 17 ± 1.4 MW to 36.1 ± 2.5 MW and the CO2flux has increased from 13.5 ± 1.7 tons per year to 36.6 ± 3.9 tons per year and the area where surface activity is present has grown, especially to the south and southeast. These changes can mostly be traced to the geothermal power plant in the area even though changes of this order of magnitude are known to be able to take place in Reykjanes without any utilization. The production involves withdrawal of large volumes of geothermal fluid which causes pressure lowering in the system. One of the consequences of the pressure lowering is the formation or increase of a boiling zone in the upper part of the system which can result in more pathways for steam towards the surface, increased heat flow and CO2 emission.
The TIR image from 2011 was calibrated from few soil measurements which were carried outat the time of the flight. The image was compared to soil temperature measurements, snowmelt tracks and a similar TIR image obtained in 2004 from Reykjanes. The comparisonrevealed a weak relationship with the results of the soil temperature measurements but these two methods showed a similar temperature distribution when visually compared. Soil temperature distribution obtained by snowmelt tracks showeda very similar distributionto the TIR2011 image. The comparison of the TIR2011 image and TIR2004 image shows that soil temperature has risen in most parts of the area and also shows an area north of the Gráa lónið lagoon which has not been included in the soil measurements in recent years where the soil temperature has risen greatly. The average soil temperature from a comparison area has risen from 7.4°C in 2004 to 10.1°C in 2011. This correlation shows that repeated TIR imagery from the same area can give reliable data to monitor changes in surface temperature.
To evaluate the effects of the grid spacing of the measurements from Reykjanes the datasets from 2009 and 2010 were split up to get two different datasets for each year. For the magnitude of temperature and CO2 flux anomalies observed in the Reykjanes geothermal area, a grid spacing of 30-50 m is not small enough to map the distribution without possible interference from randomly located measurements. Measurements carried out with a 25 m grid spacing seem to give reliable results when visually compared to the TIR image. Measurements carried out with a smaller grid spacing (17-20 m) gives a comparable distribution even though the image is more detailed. To evaluate the total CO2 flux and heat flow the measurements carried out on a 30-50 m grid spacing seem to be able to give statistically good results but that might be doubted. Measurements on a 25 m grid spacing give statistically similar results and a dataset with a smaller grid spacing does not give statistically better results or lower uncertainty.
The results from the continuous temperature measurements show a correlation between soil temperature and precipitation and the correlation is at a maximum for 96 hour precipitation indicating that precipitation lowers the soil temperature for a longer time than previously thought for Reykjanes. Temperature loggers which were located in warm soil (> 40°C) showedsudden temperature drops of tens of degrees and then rises to the previous values in a very short time (hours). These events could not be correlated to any known weather parameter but seem to depend on random micro-scale changes in heat flow.