Some Statistical Characteristics of "Geothermal Vortex" in China from 1980 to 2017

Climate prediction is of great signi�cance to the prevention of drought and �ood. The method of "Underground Synoptic" map has achieved good results in short-term climate prediction. The main content of this method is "Geothermal Vortex" (cid:0) GV(cid:0). Therefore, some characteristics of 3.2 m soil temperature anomaly (hereinafter abbreviated as T¢ 3.2 m ) and GV are analyzed by using the data of monthly soil temperature at the depth of 3.2 m from 1980 to 2017. Results show that there are regional differences in the frequency of positive anomaly. Overall, there are more positive anomalies in southern China and fewer in Western China. Soil temperature anomaly has good memory, the shortest being 5 months and the longest being nearly 2 years. Moreover, results show the nonuniform distribution of memory duration and decreases from northwest to southeast as a whole. Statistical analysis of GV shows that the average life cycle of each GV is 9.7 months, with an average horizontal scale of 604 km and a characteristic scale of 2.4 cm/s. Spatially, the GV scale is largest in Northwest, with a horizontal scale of 1087km, and the GV appear most frequently in North China. The central location of GV is uneven, mainly in the central area with strong crustal activity and large absolute vertical deformation rate or in the high value area of soil heat �ow.


Introduction
Under the background of climate change, the frequent occurrence of climate disasters and extreme weather events such as oods, droughts (Rosenzweig et al. 2001;Meehl et al. 2005) and so on in the world has caused serious economic losses (Lehner et al. 2006) and become one of the important factors affecting economic development and social progress (Guan and Zhou 2012).The impact of climate change is not only a scienti c issue, but also related to the sustainable development of society and environmental protection.China is a high incidence area of climate disasters.The departments of agriculture, water conservancy and disaster prevention and mitigation urgently need accurate short-term climate prediction information to serve national economic construction scienti cally and reduce casualties and economic losses caused by climate disasters.Therefore, short-term climate prediction has gradually become an important issue of international concern (Chou and Xu 2001).Compared with short-term weather forecasting, short-term climate prediction is an extremely di cult scienti c problem, which involves the interaction of the earth system, making the prediction mechanism very complicated."Ground-synoptic" map is a commonly used method for short-term climate prediction, which not only has a high accuracy, but also has a stable prediction skill."Ground-synoptic" map is a four-dimensional image describing the continuous evolution of various important factors in the earth's surface, in which soil temperature is considered to be the most important factor.The long memory of deep soil temperature can provide effective seasonal climate prediction signals.Tang and Gao (1995) de ned geothermal vortex using 3.2 m soil temperature and applied it to precipitation prediction.Geothermal vortex is the thermal activity unit on the "Ground-synoptic" map and the concentrated expression of the earth's internal thermal dynamic process on the earth's surface.Its analysis is an important content of "Ground-synoptic" map.
Geothermal vortex corresponds well to precipitation and earthquakes in the active unit (Jiang and Tang 1999).Tang and Gao (1997) discussed the occurrence process of xingtai earthquake in 1966 in combination with the evolution of the small-scale soil temperature eld.It was found that the merger of the four geothermal vortexes of the Xingtai earthquake and the enhancement after the merger were one of the basic causes of the earthquake.In 1998, Jiang (1998) pointed out that there was a huge geothermal vortex group activity in North China before the Tangshan earthquake, and geothermal vortex merged before the earthquake broke out, thus demonstrating the relationship between the geothermal vortex activity and the Tangshan earthquake in 1976.
The basic principle of "Ground-synoptic" map predicting short-term climate change is that a rainy area will appear in the slightly downwind direction of a geothermal vortex at a development or stability stage greater than a certain critical dimension.According to this principle, climate prediction of precipitation is transformed into prediction of the location and development stage of geothermal vortex (Tang et al.1998).Since 1984, this method has been formally applied to ood season prediction, that is, the precipitation was predicted by combining the evolution characteristics of soil temperature eld with the characteristics of earthquake eld.Based on the data of earthquake and soil temperature, Guo et al. (2004) analyzed the causes of severe ood events in the Yangtze River Basin and Nenjiang River Basin in summer 1998.It was found that the distribution of precipitation coincided well with the anomaly eld of soil temperature.The mechanism of soil temperature affecting precipitation was discussed.Adding soil temperature to the model results in continuous soil heating conditions that facilitate convective development and improve the simulation of precipitation (Liu and Avissar 1999; Xue et al. 2018).Xue et al. (2012) applied the initial conditions of cold and warm soil temperature to general circulation model operation and corresponding Eta operation, and found that the initial warm soil temperature in May in the western United States had a positive contribution to June precipitation in the southern region, while less precipitation in the northern region.Another study found that the soil temperature in northwest China in May was closely related to the East Asian summer monsoon precipitation in June and July.When the soil temperature was higher in Northwest China in May, there was more precipitation in South China in June and July.The close relationship between soil temperature in the Northwest China in May and the East Asia/Paci c teleconnection model can be used as a predictor of summer monsoon precipitation in East Asian (Wang et al. 2013).Many studies have shown that there exists an indispensable vertical heat ow between 0-3.2 m soil layer and deep soil layer (Harris and Chapman 1997;Pollack et al. 1998).In general, information in deep soils affects weather and climate over longer time scales, and information in shallow soils has a greater impact on shorter time scales (e.g., Huang et al. 1996).
The above research shows that detailed analysis of soil temperature eld and evolution characteristics of geothermal vortex is of great signi cance for short-term climate prediction.However, previous studies mostly used the 3.2 m seasonal soil temperature anomaly map to analyze the characteristics of geothermal vortexes (Tang and Gao 1995), which may not be accurate enough for the number and duration of geothermal vortexes.In order to obtain relatively accurate characteristics of geothermal vortexes, the monthly soil temperature anomaly maps were drawn using the data of 3.2 m soil temperature from January 1, 1980 to August 31, 2017, four times a day, and several characteristics of the national geothermal vortexes were statistically analyzed.

Data
The observational data used in this study are the daily soil temperatures at 3.2 m of 310 stations over China during the period of January 1, 1980, to August 31, 2017, provided by the China Meteorological Administration (Fig. 1).The dataset has eight soil layers with depths of 0, 5, 10, 20, 40, 80, 160, and 320 cm.Owing to the limitation of data availability, we only retained the stations with complete records for speci c periods and soil layers.At the same time, in order to avoid errors in the analysis, the same observation sites are used at all layers, so a total of 310 sites are screened out.We thoroughly examined the data according to the method proposed by Hu andFeng (2003, 2004b), and carried out quality control at the annual and monthly timescales.To determine abnormal and missing values, a time series of observations data were plotted at each depth of each station.In order to identify the drift in the soil temperature series, we displayed data at each site and compared the change of the target time series with those in the adjacent depth or sites, and then use regression analysis of the adjacent depth or sites to correct the drift of the soil temperature.When calculating the monthly mean value at each depth for a given station, the daily data should be less than 10 missing values, which are scattered in a given month and less than ve consecutive missing values per day.The annual mean value is calculated from the monthly mean values including all months.As seen in Fig. 1, the soil temperature observation stations are less frequent in the Northwest .

Statistical Characteristics of 3.2 m Soil Temperature Positive Anomalies
The number and duration of positive anomalies at 310 stations in China from January 1980 to August 2017 were counted (Fig. 2), i.e. the number and duration of T′ 3.2 m > 0°C at each station.The duration of soil temperature anomaly is the continuous positive or negative duration of soil temperature anomaly.According to the statistical results of the occurrence times of positive anomalies at each station, the contour map as shown in Fig. 3 is drawn.From the gure, it can be seen that there are regional differences in the number of positive anomalies occurring in the 38 years.Overall, there are more positive anomalies in southern China and fewer in Western China.Among them, there are several obvious high-value areas: the positive anomaly in southern Guangxi occurs most frequently, reaching 46 times.Guizhou province, central and eastern Sichuan province and southeastern China take the second place, about 34-42 times.About 30-38 times in southeastern Liaoning province and Northern Hebei province.There are about 30 times at the junction of eastern Qinghai and Gansu province.At the same time, there are also relatively obvious low-value areas: The northwest China is generally less than 18 times.The number of positive anomaly western Hubei, northern Ningxia and southeastern Gansu are less than 10 times.
Similar to soil moisture, soil temperature has the ability to memorize atmospheric anomalies (Yang and Zhang 2015).
However, the memory duration of soil temperature anomalies in China remains unclear.Therefore, this paper presents the spatial difference of the memory duration of soil temperature anomaly at 3.2 m (Fig. 4).On the whole, 3.2 m soil temperature anomaly lasted for a long time, the shortest lasted for 5 months and the longest lasted for nearly two years.And the memory duration has signi cant spatial distribution characteristics: It decreases from northwest to southeast.In Northwest China, the memory duration of 3.2 m soil temperature is about 12-19 months.The soil temperature memory duration in the southeastern is 4-9 months.The memory duration of soil temperature in Northeast China is about 9-15 months.
In addition, the statistical results showed that there were 2431 positive anomalies in China during the 38 years.The average memory duration of each positive anomaly was 9.9 months, about 0.83a.Figure 5 shows the relationship between the duration of soil temperature anomaly and the number of occurrences.The largest number of normal anomalies were observed within one month, and 85% of them were observed within one year.With the increase of duration, the number of positive anomalies decreases.

Some statistical characteristics of "Geothermal Vortex"
The key to predicting short-term climate change by "Ground-synoptic" map is to determine the location and movement of geothermal vortex.Therefore, some characteristics of geothermal vortex are mainly analyzed in this section.Based on the 3.2 m monthly anomaly data from January 1980 to August 2017, 452 monthly anomaly maps were drawn.Contours are drawn on each 3.2 m monthly anomaly map of soil temperature, and the high temperature center with soil temperature anomaly T′ 3.2 m ≥ 0.5 °C could be analyzed.At least part of the heat generated by this high temperature zone comes from deep underground.When the disturbance of monthly soil temperature anomaly reaches 3.2 m, the amplitude of soil temperature will decrease to 0.05-0.15°C.The high temperature area often accompanied by earthquakes is de ned as "Geothermal Area".We stipulate that on the T-map (Fig. 6), there are patches of T′ 3.2 m > 0°C areas, in which at least one station T′ 3.2 m ≥ 0.5 °C is surrounded by the low temperature zone T′ 3.2 m < 0°C, and the area T′ 3.2 m > 0°C is called "Geothermal Vortex".For example, the northwest region in Fig. 6.
As for the horizontal scale L of " Geothermal Vortex ", it is stipulated in this paper that since the shape of each "geothermal vortex" is approximately elliptic, as shown in Fig. 6 in Northwest China, the lengths of long axis (H) and short axis (S) can be calculated according to the longitude and latitude coordinates of the edge, then L= (H + S)/2 is the horizontal scale of " Geothermal Vortex ".After these speci c de nitions of "Geothermal Vortex" are given, its spatial characteristics can be statistically analyzed.1 shows the statistics of the maintenance time of "Geothermal Vortex".It can be seen that the most cases of "Geothermal Vortex" maintenance are 4-7 months, followed by 1-3 months.The number of cases of "Geothermal Vortex" maintenance for more than 1.5 years is very small, and the longest is more than 13.4 years.The average duration of each geothermal vortex is 9.7 months.The average number of "Geothermal Vortex" on each map is 5.The horizontal scale of a "Geothermal Vortex" varies at different stages of its development.The "Geothermal Vortex" accompanying the earthquake is small in horizontal scale when it is formed at the beginning, and the largest when the earthquake occurs, and then decreases (Tang and Gao 1995).In this paper, a graph with the largest range of each "Geothermal Vortex" was taken to make statistics on the horizontal scale.Table 2 shows that the largest number of horizontal scale is 400-500km, followed by 500-600km.The number of "Geothermal Vortex" with the horizontal scale of 300-400km, 600-800km and 1000-1888km was the same, and the smallest number was less than 200km.The largest horizontal scale was 1888km, which appeared in Xinjiang, Qinghai and Tibet in August 2000.
From Table 2, the average horizontal scale of 238 geothermal vortexes is 604 km, which is close to the depth of the upper mantle (670 km).If the thickness of the lithosphere (70-80 km) is subtracted, it can be concluded that the thickness of the upper mantle is basically equal to the average horizontal scale of the geothermal vortex.The characteristic velocity of the "Geothermal Vortex" (V = L/t ≈ 746 km/a ≈ 2.4 cm/s) can be obtained from its average duration (t = 0.81a) and average horizontal scale (L = 604 km).(Tang and Gao 1995).At the same time, it can be clearly seen that the center of the "Geothermal Vortex" moves horizontally from east to west or from north to south.However, most of the centers move eastward and tend to move toward higher latitudes, which is similar to the movement of atmospheric cyclones.As time goes by, one "GV" may split into two, or two "GV" may merge into one.The evolution of "GV" is complicated, and all "GV" moving on the same track can be called a "GV" family, with a life history of roughly 10 years.The vortex moves about 200-500km a year, almost as fast as a single one.Meanwhile, the path of geothermal vortex movement is similar to that of continental cyclone eastward moving out of the sea, mostly from the Yangtze Estuary to the East China Sea, from North China to the Yellow Sea or northeast to the east.Low-value systems in the atmosphere tend to form over geothermal vortexes (Tang and Gao 1995).

Conclusions and discussion
Based on 3.2 m soil temperature data from 310 stations in China, some characteristics of soil temperature anomaly and "Geothermal Vortex" are analyzed.Generally speaking, there were more positive anomalies in the south and fewer in the west over the past 38 years.The anomalous duration of 3.2 m soil temperature is longer, the shortest is ve months, the longest is nearly two years, and the duration decreases from northwest to southeast.
"Geothermal Vortex" is the thermal activity unit on "Ground-synoptic" map.The location and movement of "Geothermal Vortex" should be determined rst to predict short-term climate change by using "Ground-synoptic" map.In this paper, 452 monthly anomaly maps are used to calculate the duration and horizontal scale of geothermal vortex.In 38 years, 238 "Geothermal Vortex" appeared, each of which had an average life cycle of 9.7 months, an average horizontal scale of 604km, and a characteristic velocity of about 2.4cm/s.The occurrence time and end time of each "Geothermal Vortex" have distinct seasonal characteristics.The scale of "Geothermal Vortex" in Northwest China is the largest, and the "Geothermal Vortex" appears most in North China.The central position of "Geothermal Vortex" is uneven, which mainly occurs in the central area where crustal activity is intense and the absolute value of vertical deformation rate is large, or in the high value area of soil heat ow.The moving path of the "Geothermal Vortex" center is similar to that of the cyclone.The Average memory duration of soil temperature anomaly at 3.2 m (Unit: month) The relationship between the number of positive anomalies and their memory duration

However, there are
still a small number of Geothermal vortexes whose central position and moving path are di cult to track.Declarations Funding This was supported by the Second Tibetan Plateau Scienti c Expedition and Research Program (STEP), Ministry of Science and Technology, China (2019QZKK010303) and Shandong Provincial Natural Science Foundation (ZR2021YQ28) and the open project of Key Laboratory of Land Surface Process and Climate Change in Cold and Arid Regions, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences "Study on the Change of Soil Thermal condition of Qinghai-Tibetan Plateau and Its Impact on Climate Change".

Figures Figure 1
Figures

Table 3
Average horizontal scale of "Geothermal Vortex" in different regions (Unit: km).

Table 3
shows the statistical characteristics of geothermal vortexes in different regions.It can be seen that the characteristics of geothermal vortexes in different regions are different during the development of geothermal vortexes.The scale of geothermal vortexes in Northwest China is the largest, followed by Southern China, with a horizontal scale of about 560 km, Eastern China of about 523 km and Northern China of about 460 km.Since there is a basically xed "Geothermal Vortex" center at the junction of southeastern Gansu and Shaanxi province, this region is considered as a separate region.In addition, the largest number of cases occurred in North China, accounting for 41% of the total geothermal vortexes, followed by the Qinghai-Tibet Plateau and Xinjiang, accounting for 18.5%, and the remaining region accounted for 40.5%.Figure7records the coordinates of the central points of 2353 geothermal vortexes on 452 maps.Different colors indicate the number of times that the central point of the "Geothermal Vortex" appears at this position.It can be seen that the distribution of the "Geothermal Vortex" is not uniform.It is concentrated in some areas, namely, northern Xinjiang, central Tibet, Eastern Gansu and Qinghai, northeastern China, central North China, southeastern Gansu, the middle reaches of the Yangtze River, Yunnan-Guizhou Plateau, and coastal areas of Fujian-Guangdong.Most of these areas are areas with strong crustal activity or high values of soil heat ow