GIS Based Identification Model For The Assessment Of Shallow Geothermal Potential

Energy science have been not only greatly promoting the social progress but also providing the human beings with unprecedented convenience and comfortable life.Over the past century,however,the characteristics of the fossil fuel energy structure has brought the serious negative impact to the environment.The energy crisis and the increasingly serious environmental problems have been restricting the sustainable development of global economy and the society.In order to achieve a rapid economic development society,the use of renewable energy has become the consensus of the whole society.In this case,geothermal energy has emerged with the advantages of green,continuous and stable utilization has broad development prospect.As a major energy consuming country,China's geothermal energy reserves were large,the distribution of geothermal potential was wide and the development prospect was broad and the market potential was huge.Although the abundant geothermal resources have not been exploited,geothermal energy has a great development potential.However,the exploration and development still face certain challenges.For hydro-thermal geothermal energy,the initial exploration cost was up to 50% of the total investment in the whole project,and thegeothermal drilling risk in the exploration phase is large where the success rate for drilling was only 25%.At the same time,geothermal resources evaluation and exploration model are too simple.Due to the lack of detailed exploration in China's key areas,geothermal resources exploration cannot really serve the construction of geothermal projects.Therefore,it is important to establish a reliable prediction model to improve the accuracy of geothermal target area.With the probabilistic statistical model based on GIS and the evaluation of method,the resource potential and geological hazard assessment were widely applied and developed.Hence,by evaluating the influence factors of geothermal distribution,the prediction of Tengchong geothermal potential area based on mathematical foundation will be feasible.Take Tengchong area as an example,the study area(24 ° 00 '00-26 ° N,98 °00'-99 ° 00 'E)including Tengchong County,part of Lianghe County,Yingjiang County,and Dehong prefecture in northeastern of Yunnan Province.The regional tectonics is related to the subduction and collision of the Indian plate and Eurasian plate.The underground fracture and tectonic basin are developed and the magmatic activity is very intense.The study area belongs to the Himalayan earthquake belt of the Burma,filled with seismic tectonic system,earthquakes,volcanoes,hot springs and other complex geological tectonic characteristics,has been one of the most intense regional geothermal activity during the late Cenozoic volcanic activities.The geophysical data showed that there were multiple volcanic magma capsule,high heat flux,low gravity and magnetic anomaly characteristics:deep circulation of groundwater bring the water to near surface and heated from magma capsule.In this case,geothermal anomaly can be found.In Tengchong region,western Yunnan province,curtain shell heat flux ratio is 1.53 indicating a very high heat flow value.Negative gravity anomaly caused by material loss and by volcanic eruption of magma material can be found int this area.The remaining magnetism was obtained from the cooling process after the hot magma overflow.In addition,several eruptions have brought the study area with plenty of magnetic irregularities.So far,there have been frequent seismic activity in Tengchong and most of them are in the hot sea area,south of Tengchong County.Due to the well developed seismic tectonic fracture,moho deep fracture exists,where the main fractures located in Tengchong region are Gaoligong fault,Nujiang fault,Dayingjiang fault and Longling fracture.The collected preliminary data and analysis showed that the geothermal anomaly in the study area was related to the distribution and quantity and terrestrial heat flow,magmatic rocks,fault distribution,earthquake distribution,bouguer gravity anomaly,magnetic anomaly distribution and rivers.Based on the GIS platform,geothermal anomaly area can be identified according to relationship between the number and location of hot-springs and their surrounding geological and geophysical conditions.In this paper,the idea that through mathematical probability model is aim to assess the geothermal potential regions where has not been developed or failed to fully exploited.Driven by knowledge of mathematical probability method including the amountof information model,information entropy based weighted information model,certainty factor model,index overlay based certainty factor model,information entropy based weighted certainty factor model.Although many models have significant effect on the evaluation and application of geothermal potential area,the above models are not rarely used in the evaluation and application of geothermal potential area in Tengchong region.The available publicly data sources used for analysis were put forward as follows: epicentre data,fault distribution,Bouguer gravity anomalies,Landsat7 ETM + image data,magnetic anomalies data and digital elevation model data.Before modeling application,the original data were extracted as Gutenberg-Richter B value map,distance to fault map,distance to the main grabens map,land surface temperature map,magnetic anomalies map and distance to rivers map respectively.The extracted impact factor maps has obvious spatial relationship with the geothermal distribution,where the higher B value map contains most of geothermal training points,and most geothermal training points are close to the fault distribution,the higher surface temperature and a closer to the rivers indicate more training points present.Due to the overall value is low in the magnetic anomaly map,there is no obvious relationship with the geothermal distribution.It is important to note that the above method used have a strict assumption that the impact factors are conditionally independent between each impact pactor map pairs.However,the impact factor data is not easy to obtain,and even if sufficient basic data is obtained,a effective algorithm evaluation method is stillneeded.Therefore,such problem for testing independence of the factors was often ignored before modeling application.In this paper,KMO and Bartlett's sphericity test and factor analysis were used to test the conditional independence between impact factor map pairs,and the factors with a closer relations with other maps were excluded to ensure the reliability of the models.In the application for the series of information models,the results showed that the correlation between the distance to main grabens map and the magnetic anomalies map and the LST map were obvious,thus the distance to main grabens map was excluded.Other impact factors were confirmed conditionally independent and used to evaluate geothermal potential regions.At the same time,in the model application for the series of certainty factor models,the factor analysis results show that the five used impact factors are conditionally independent and can be used in model application.Finally,five prediction maps were generated where most of geothermal training points were distributed in favorable geothermal areas,indicating that potential geothermal areas have been fully developed and utilized,and demonstrating the effectiveness of the proposed prediction models.Take information mode as an example,four geothermal potential regions that have not been explored were selected from classified favorability map including the area distributed along the Nujiang river,the basin along the Longchuan river,the area along the Nandi river,and the area between the Mingguang river and Zhimeishan river.The total area of geothermal potential area that has not beendeveloped or exploited is 1835.6 km2,accounting for 6.72% of the whole research area.Through the Kappa coefficient analysis,the index analysis and area ratio analysis,the consistency and accuracy of the five models can be presented.According to the results of Kappa coefficient analysis,the five models were considered to be basically consistent with the index range form 0.664 to 0.744.From the results of index analysis,the results show that the information model is more accurate than the weighted information model.For the series certainty factor models,the index overlay based certainty factor model has a highest accuracy result.In the area ratio analysis,the comparison results of the series information model are consistent with the success index result.But the modified certainty factor model has the most accurate prediction accuracy.Generally,through the successful index analysis and area ratio analysis results,all of the prediction models can be well used in evaluation of geothermal potential regions.However,how to chose a suitable forecasting model is not easy to decided.In the paper's point of view,the assessment result indicating that the way to choose a better method is depended by the user's specific circumstances.If someone need a more accuracy in assessment in Tengchong area,the information model and modified certainty factor model will be a choice.If the convenience is the fist need to be considered,it is suggested that index overlay based certainty factor model will be the first choice.If someone need a more reasonable result the weighted information model and the weighted certainty factor model will be thefirst choice.