Theoretical method of metallogenic prognosis

First, the theoretical basis of metallogenic prognosis

Metallogenic prediction is to comprehensively study geological prospecting information such as geology, geophysics, geochemistry and remote sensing geology, analyze metallogenic geological conditions, summarize metallogenic regularity, establish metallogenic model, apply the principle of "from known to unknown" to evaluate resources in unknown areas or delineate different levels of prediction areas, propose key exploration sections or arrange specific exploration projects, so as to improve the scientificity and effectiveness of prospecting work and improve the degree of metallogenic geological research (Zhao).

1. Abundance theory of crustal mineral resources

The main point of this theory is that the distribution of elements in the crust is uneven, and the local enrichment of elements forms mineral resources with economic value. There is no region in the crust where there are no mineral resources or resources are completely exhausted, and there is no region where resources are completely concentrated (Zhu Yusheng, 1984). Resource evaluation is to determine what kind of resources exist in a certain area and how many resources there are. This theory clarifies the objective fact that mineral resources exist in the earth's crust and the possibility of evaluating them.

2. Similar analogy theory

Similar geological environments should have similar deposits. This is the theoretical guiding principle for establishing the quantitative relationship between mineral resources and geological environment. Under the guidance of this theoretical principle, the method of "from known to unknown" is adopted to predict mineral resources, that is, the evaluation model of the relationship between mineral resources and geological conditions is established in the known area, and it is extrapolated to the prediction area with similar geological structural conditions in the known area to estimate the resource quantity in the prediction area (Zhu Yusheng, 1984).

3. The theory of mineral resources prediction model, stochastic function theory of mineralization and functional relationship between ore-controlling factors and mineralization.

This is one of the achievements in the field of mathematical geology research. The mineral deposit model is directly or indirectly used in the prediction and evaluation of mineral resources. In practical work, a mathematical model for evaluating the relationship between mineral resources and geological conditions is established by combining geological data and experience, and mineral resources are predicted according to the model. Geological theory is the basis of establishing mathematical model of mineral resources evaluation (Zhu Yusheng, 1984).

4. Theory of synthesis and decomposition of geological variables

Geological variables are the basis of establishing the prediction and evaluation model of mineral resources. The geological variables related to mineral resources are selected from various original geological data, and the prediction and evaluation model of mineral resources is established, and the comprehensive information is used to predict and evaluate mineral resources. This is the meaning of data synthesis. Modern popular comprehensive information metallogenic prediction and comprehensive geological information prediction technology of deposit model are the deepening development of comprehensive theory of geological variables. On the premise of geological theory, comprehensive information metallogenic prediction is to study multiple information such as geology, geophysics, geochemistry and remote sensing from the perspective of geological evolution, and then establish comprehensive information prospecting model and comprehensive information prediction model, and systematically evaluate the research area with comprehensive information prediction model as a tool.

For a certain kind of geological variables, the data used for evaluation has gone through a long geological period, which is the synthesis of its geological and historical behavior; For the same time process, this variable can be regarded as the synthesis of several more local different geological processes. Based on the geological variables representing the final results of a series of geological processes, this paper analyzes their geological action behaviors in different time and space processes, especially those related to the origin of resources, and predicts the types, locations or quantities of mineral resources. This is the significance of variable decomposition in mineral resources prediction and evaluation (Zhu Yusheng, 1984).

5. Metallogenic series theory

The comprehensive discussion of the concept of ore-forming series of deposits was put forward by Chinese geologists in the process of long-term prospecting and geological research of deposits. It will study a group of mineral deposits related to geological mineralization in a region and their relationship in space, time and genesis as a whole. This is of great significance to deeply understand the metallogenic regularity and guide the exploration of mineral deposits. As far as prospecting in a certain region is concerned, on the basis of studying the regional geological structure background in detail, the concept of metallogenic series can be used to comprehensively analyze and understand the metallogenic environment, ore-controlling factors, mineralization and possible deposit types in this region, that is, to establish an overall concept and look for unknown deposits according to known deposits, so as to expand the prospecting ideas and clarify the prospecting direction.

6. Geological anomaly theory

Geological anomalies refer to geological bodies or combinations of geological bodies that are obviously different from the surrounding environment in composition, structure, structure and genetic sequence. If the numerical value (or numerical interval) is used as the threshold to represent the background field, any field that exceeds or falls below the threshold constitutes a geological anomaly. Geological anomalies often show differences in geophysical field, geochemical field and remote sensing image anomalies, and are often comprehensive anomalies. Geological anomalies of different scales not only have different delineation marks and different degrees of scope and size characteristics, but also have different relations with mineralization. Global geological anomalies are crustal structural anomalies, regional geological anomalies are geological anomalies that control the inter-provincial distribution of metallogenic belts, provinces and metallogenic areas, and local geological anomalies are geological anomalies that control the output of ore fields, deposits and ore bodies in metallogenic areas (Zhao et al., 2006).

7. Inertia principle

The principle of inertia refers to the continuity of objective things in the process of development and change. The inertia phenomenon of ore-forming events and their products-ore deposits presents a stable trend in time and space. The more stable this trend is, the stronger the inertia is, and the less likely it is to be disturbed by external factors and change its own trend (Zhao et al., 2006). For example, the scale and extension direction of some large metallogenic belts and vein ore bodies are generally stable. The trend extrapolation method commonly used in metallogenic prediction is developed based on the spatial inertia phenomenon of the related characteristics of geological bodies.

8. Relevant principles

The principle of correlation means that the occurrence and change of any ore-forming event are not isolated, but developed under the interaction with other geological processes, and this interaction often shows causality (Zhao et al., 2006). For example, the target mineral resources of metallogenic prediction are usually closely related to various rocks and structures, and some types of deposits are special products of specific geological processes. The correlation principle is helpful for us to comprehensively and deeply analyze various geological factors related to mineralization, so as to correctly understand the relevant characteristics of the deposit, summarize the metallogenic regularity, and then make correct predictions.

9. Geological interpretation theory

Geological interpretation is to transform the evaluation model into the concept of geological genesis and resource characteristics (estimated deposit quantity, tonnage or grade) (Zhu Yusheng, 1984). The key point is to supplement the mineral resources information not included in the established mineral resources evaluation model with the geological theory and accumulated experience mastered by geologists, and transform it into the concepts of geology and resources.

Second, the main methods of metallogenic prognosis

(A) the basic principles and characteristics of metallogenic prognosis

1. The principle from known to unknown

The prediction of mineral resources in unknown areas is often based on the model established in known areas to evaluate the resources in unknown areas. Therefore, the geological structure conditions of the unknown area are highly similar to those of the known area. This is actually a concrete application of analogy theory.

2. Establish the quantitative relationship between mineral resources and geological conditions.

This is a necessary condition for establishing an evaluation model of mineral resources, which plays a decisive role in the prediction and evaluation of unknown resources and is a difficult link in the prediction and evaluation work. On the surface, some prediction and evaluation models only study the distribution and change of data parameters, without involving geological conditions, but in fact, this distribution and change is dominated by geological conditions, which implies the role of geological conditions (Zhu Yusheng, 1984).

3. The knowledge and experience of geological experts affect the prediction and evaluation of mineral resources.

Some evaluation models are based on the knowledge and experience of geological experts, but in fact they are also based on the relationship between mineral resources and various geological conditions, which are implicit in the experience and knowledge of geological experts (Zhu Yusheng, 1984). In this case, the knowledge and experience of geological experts play a decisive role in the prediction and evaluation of mineral resources, which requires comprehensive research and demonstration by high-level experts from different majors.

4. As rich as possible input information and as simple as possible evaluation results.

The prediction and evaluation of mineral resources should use useful geological information as much as possible to ensure the accuracy of the prediction results. But to sum up, it should be as simple as possible, which is convenient for geologists to identify and relevant departments to apply.

5. The result of quantitative estimation of mineral resources is probabilistic.

Due to the complexity of mineralization, our geological knowledge is far from enough to summarize an accurate mathematical model for prediction and evaluation. Most of the evaluation models of mineral resources established by us are random, and the corresponding mineral resources predicted are also random (Zhu Yusheng, 1984). Therefore, the predicted mineral resources are probabilistic, that is to say, the estimated mineral resources are not absolute, but judged in a certain probability sense.

6. The principle of minimum risk and maximum ore content

The prediction results required to be submitted should delineate the spatial position of the prospecting target area with the smallest area on the premise of the minimum possibility of missing concealed deposits.

7. Optimization evaluation principle

Optimization evaluation refers to that the forecaster consciously intervenes in the composition of the model according to his own understanding of the metallogenic regularity and ore-controlling factors, and makes directional transformation of the model which is beneficial to mineralization (or strengthens the metallogenic information) (but on the premise of not changing the prediction target of the model), so that the model highlights some important prediction indicators (or ore-controlling factors) information, suppresses some information which is of little significance or strong interference to mineralization, and forces the model to concentrate information in the favorable direction of mineralization, highlight the prospecting criteria and gradually approach the potential.

(2) Brief introduction of metallogenic prediction and evaluation methods.

Metallogenic prediction is to estimate or infer the unknown characteristics of past metallogenic events. The process of prediction is a rigorous scientific logical thinking process, including observation, analysis, induction and reasoning (Zhao et al., 2006). There are dozens of specific metallogenic prediction methods, which can be divided into three categories according to the different scope of metallogenic prediction and evaluation: regional mineral resources prediction and evaluation, mining area prediction and evaluation and deposit prediction and evaluation, and the specific methods adopted in each category are different (Zhu Yusheng, 1984).

1. Prediction and evaluation method of regional mineral resources.

The evaluation methods of (1) non-geological indicators include Sipov's law, historical yield method, Lasky's law, Hewitt curve, spatial distribution statistical model, etc.

(2) Subjective evaluation methods include geological analogy method, simple subjective probability method, complex subjective probability method, subjective network method and Delphi method.

(3) Simple geological marker model evaluation methods, such as volume estimation method, regional value evaluation method, trend surface analysis method, abundance estimation method, etc.

(4) Evaluation methods of qualitative geological marker model, such as fuzzy mathematics, logical information method, feature analysis method, quantitative theory, probability regression, rank correlation analysis method, Monte Carlo method, etc.

2. Prediction and evaluation methods of total mineral resources in mining areas

(1) The subjective evaluation method is the same as the regional evaluation method (2).

(2) Evaluation models of metallogenic indicators, such as judgment analysis, cluster analysis, regression analysis, factor analysis, correspondence analysis, deposit model method, genetic geological model method, etc.

(3) The evaluation model of qualitative metallogenic geological indicators is the same as the regional evaluation method (4).

(4) Extrapolation of trend, including extrapolating external characteristics of ore bodies, extrapolating internal characteristics of ore bodies, extrapolating metallogenic conditions, extrapolating ore-controlling factors, extrapolating prediction indexes, extrapolating metallogenic regularity, etc. Zhao et al., 2006.

3. Prediction method of total mineral resources of the deposit

(1) Geometrical Method.

(2) Geological and geochemical methods.

(3) Geological-geophysical methods.

④ Trend extrapolation method is the same as mining area evaluation ④.

The prediction and evaluation methods of mineral resources in different regions are relative, and various methods can be flexibly selected in specific prediction and evaluation. Geological analogy is the real basis of various mineral resources prediction and evaluation methods. Modern mineral resources prediction and evaluation methods are not only related to traditional geological methods, but also developed quantitative evaluation methods. On the basis of geological research, various models are established around the overall goal of mineral resources prediction and evaluation, and the potential resources of a geographical area, metallogenic belt or smaller area (ore deposit) are estimated by mathematical methods.

Three. Methods used for gold deposit prediction in this project.

(1) Quantitative prediction method of total gold deposits in northwest Jiao Jiao.

The geological conditions of gold deposits in the northwest of Jiaodong are complex and the prospecting information is diverse, so it is difficult to correctly evaluate their total resources with a single and simple prediction method. On the basis of predecessors' work, this work uses a variety of prediction and evaluation methods based on comprehensive information metallogenic prediction to predict the total amount of gold resources in the northwest of Jiaodong.

1. Comprehensive information metallogenic prediction

By using mathematical geological methods and computer, this paper comprehensively explains all kinds of prospecting information related to minerals, such as geological elements, geophysical and geochemical exploration, heavy sand anomalies and so on. Comprehensive information metallogenic prediction emphasizes geological premise, extracts comprehensive information as unit to establish comprehensive information model, and predicts minerals by analogy. Based on the study of typical mineral deposits, regional metallogenic regularity and metallogenic conditions, this prediction extracts useful information related to mineralization, makes statistical comparison between information and gold resources, and determines the correlation between useful information and gold resources. Based on the analysis of useful information, the geological variables are selected and assigned, and the variables are divided into fixed variables and quantitative variables.

(1) Positioning variable selection.

The selection of positioning variables mainly considers the relationship between useful information and resource characteristics, whether there are statistical laws in the unit and the nature of information. In order to realize the location prediction of mineral resources, a three-state and two-state variable system is established, and variables are taken from nine aspects of information such as stratum, structure, magmatic rock, gravity, aeromagnetism, geophysical inference, heavy sand, geochemical exploration and remote sensing.

The binary variable system * * * selected 49 variables:

Strata: ① Archean metamorphic rock series; ② Jingshan Group and Fenzishan Group; (3) The strata are exposed in slices; (4) The stratum is exposed in the form of debris.

Structure: ⑤ The main structure is Grade II structure; ⑥ The main structure is Grade III structure; ⑦ The main tectonic directions are NE direction and NNE direction; (8) secondary structure development; Pet-name ruby structural fracture zone development; Attending ductile deformation development; ? The secondary structure passes through the middle of the unit; Hey? The swallow unit is located in the lower wall of the secondary structure; Answer? The structural alteration zone is completely zoned; Hey? The alteration types are sericitization, silicification and pyritization.

Magmatic rocks: Archean TTG rock series and Jurassic Linglong granite (Jiuqu, Yunshan and Cuizhao rock bodies); ? Cretaceous granites of guojialing, Wendeng and Weideshan; ? Rock mass facies belt is marginal facies; Hey? Gneiss and porphyritic medium-coarse grained granite; Ying Ying? The contact zone is fault contact; Hey? The contact zone is intrusive contact; Hey? Seasonal dikes, lamprophyres and diabase porphyrite dikes are developed.

Gravity:? The contour lines are gentle slope zone, nose zone and twisted zone, and the speed is less than1.5×10-5m/(S2 km). ? The contour line of Zhen A is a gentle gradient belt with curvature, and the speed is (1.5 ~ 2.5) ×10-5m/(S2 km). Answer? The gravity field value is between 0 ~ 30× 10-5m/S2.

Magnetic field: low alternating field, low positive field and low negative field; Hey? The magNEtic field axis has two directions: ne and NNE.

Geophysical inference: what? East-west basement structure ≥10 km; ? East-west basement structure 5km, huh? Yingdong and NNE structures are > 5 km; Hey? 3km structure in NE and NNE direction, huh? North-north-east, north-east and near-east-west structures meet; Hey? Overlapping, harbor-shaped and tongue-shaped parts of rock mass; Answer? There are concealed rock masses.

Heavy sand:? Grade A Ⅰ and Ⅱ abnormal gold and heavy sand; ? The combination of grade ⅰ and ⅱ, which is dominated by gold, is abnormal; ? The heavy sand anomaly is in good agreement with the structure; ? Abnormal scale (ratio to unit area) is greater than 50%.

Geochemical exploration:? Anomalous assemblage of geochemical exploration; ? Yan Ying's other combinations are not normal; Hey? The scale of geochemical anomalies (ratio to unit area) is more than 50%; Hey? Geochemical anomalies are in good agreement with structures; ? The abnormal value of gold is > 4×10-9; ? Abnormal value of gold (2 ~ 4) × 10-9.

Remote sensing: what? Development of ring structure; Hey? Ring structure exists; Hey? Development of linear structure for electronic warfare:? Linear structure develops in other directions; Hey? Loop intersection and complicated cutting degree; ? The crossing and cutting degree of Yu Ying Ring Road is simple.

Three-state variable system * * * selected 3 1 geological variables. Their names and their relationship with mineralization are shown in Table 9- 1.

(2) Selection of quantitative variables.

The quantitative forecast variable is a descriptive quantitative variable, which can indicate the difference of forecast object scale and reflect the scale level of resources. At the same time, these variables are also continuous variables, and the resources of geological units are predicted by regression prediction model.

Descriptive variables * * * include 7 items and 25 variables:

1) Predict the distance between the unit and the secondary fault: ① With the increase of the distance, the resource scale decreases, but there is no obvious linear relationship between them; (2) Super-large and large deposit units are mostly located near the fault zone; Units with a small amount of sedimentation are more than 4 kilometers away from the main fault; The unit regularity of medium-sized deposits is not obvious, but it can be near or far. It shows that this information has the ability to distinguish large and small mines. Therefore, the structure: ① 0 km; ②< 4km； ③ Three variables ≥ 4 km.

Table 9- 1 list of three-state variable types

2) Distance from the unit to the structural intersection: As the distance from the structural intersection increases, the unit deposition scale tends to decrease. Most of the units bearing super-large deposits are at the intersection, most of the units bearing small deposits are in the range above the intersection 10km, and most of the units bearing medium-sized deposits are in the range of 5 ~ 10 km. According to this setting ① < 5 km; ②5 ~ 10km； ③ Three variables > 10 km.

3) Width of ore-controlling fault zone: The scale of gold reserves in northwest Jiaodong increases with the increase of the width of fault zone. The width of ore-controlling fault zone in small and medium-sized deposit units is mostly less than 100 m, and that in super-large deposit units is more than100 m. The width of ore-controlling fault zone in large deposit units varies greatly. Based on this, it is determined that ① ≥ 50m; ②50 ~ 10m； ③ Three variables < 10m.

4) Ratio of abnormal area to unit area of gold dispersed flow: There is a certain linear relationship between unit deposit size and the ratio of abnormal area to unit area of gold dispersed flow, only a few units swing greatly, and units of different sizes are relatively concentrated. Can be divided into ① > 90%; ②70%~90%; ③25%~70%; (4) < 25%, that is, four variables.

5) gold anomaly area: ① > 70km2; ②30 ~ 70 km2； ③ 10 ~ 30k m2； ④ Four types < 10km2, constructed as four variables.

6) Abnormal concentration of gold: ① > 200×10-9; ②(50~200)× 10-9; ③(20~50)× 10-9; ④ Four levels are less than 20× 10-9, and are constructed as four variables.

7) Ratio of area productivity to unit area: ① > 200; ② 100~200; ③ 10~ 100; ④ The four ratio intervals of < 10 are constructed as four variables.

2. Evaluation method of qualitative geological marker model

The specific method type used in the prediction and evaluation of the total gold resources in the northwest of Jiaodong is the qualitative geological index model evaluation method in the prediction of the total regional mineral resources. On the basis of geological unit division and variable extraction, the model unit is established, and through the study of the model unit, the mathematical model is established to quantitatively predict the ore field position and optimize the prospecting target area. It involves four mathematical prediction and evaluation methods: feature analysis-judging the distribution position of gold resources, logical information method-evaluating the scale level of resources, Monte Carlo method-predicting the amount of resources in metallogenic belt (field), and regression analysis method-determining the spatial distribution of resources.

Feature analysis, also known as decision simulation, is a mathematical geological method used for statistical prediction of mineral resources. Its principle is to extract comprehensive features from multiple mineral statistical prediction variables, and establish the quantitative relationship between simulation area and prediction area according to the comprehensive features, so as to achieve the purpose of predicting unknown areas. Because the calculation methods used are different, the models of feature analysis are also different. Three commonly used models are product matrix vector length model, product matrix principal component model and probability matrix principal component model. Feature analysis can help us reduce the uncertainty of resource evaluation results caused by incomplete original data. It applies the three-dimensional environment of the deposit (including geological environment, physical characteristics, chemical characteristics and satellite image characteristics) and the data of the genesis and formation of the deposit (that is, the genesis) to establish, test and apply the deposit model, and quickly determine the similarity between the evaluation object (unit or occurrence) in the evaluation area and the known model, or the favorable degree of producing the deposit.

Logical information method is one of the methods to evaluate mineral resources by using qualitative geological data. This method is a comprehensive mathematical analysis method based on mathematical logic, combinatorial analysis and probability statistics. By means of combinatorial analysis and logical operation, the similarity of the structural relationship of the observed object is compared, and the role of a single element in the structure is determined. The essence of logical information method is to compare and predict the structural changes and structural similarities of observed data. Logical information method is an effective method to predict the scale of resources. By reasonably grading the known model ore field units, the variation sequence screening variables are established, and the symbol weight, model weight and object weight of the prediction unit are calculated to predict the resource scale.

Monte Carlo method, also known as statistical experiment method and random simulation method, is a method to solve approximate solutions of mathematical problems through statistical experiments and random simulation of random variables. It is a stochastic simulation method widely used in geological problems. Monte Carlo simulation resources can be roughly divided into the following processes: ① building a probability model, that is, establishing the relationship between resources and parameters; ② Establish the statistical distribution of parameters; (3) generating random numbers; ④ Sampling to form resource distribution; ⑤ Use the resource distribution model to estimate the resources in the forecast area, and then make an evaluation.

Regression analysis, also known as factor analysis, is one of the most commonly used forecasting methods in economic forecasting. Find out the mathematical relationship between an economic variable and some variables (explanatory variables) that are regarded as the main reasons for the change, that is, establish a mathematical model, and then give the future values of exogenous variables (that is, variables that are less affected by the model and determined by external conditions) in some way, bring these values into the mathematical model, and calculate the future values of the economic variables to be predicted, that is, the predicted values. This method is also widely used in mineral resources evaluation. The main reasons are: firstly, it can not only study the relationship between variables, but also estimate the value of another variable (dependent variable) according to one or several variable values (independent variables) and infer the relationship between variables; Second, we can find the main and secondary independent variables that affect the dependent variables and determine the relationship between these variables; Thirdly, stepwise regression in regression analysis can automatically select a group of independent variables that are "closest" to the dependent variable from a large number of optional independent variables, establish an evaluation model of the relationship between resources and geological conditions, and estimate the resources in the prediction area more directly. There are many mathematical models of regression analysis, mainly including: unitary linear regression, multiple linear regression, stepwise regression, principal component regression, nonlinear regression, event probability regression, partial correlation and multiple regression, ridge regression, typical regression analysis and multiple regression.

(2) Prediction method of deep gold deposits in Jiaojia belt.

The prediction of deep gold deposits in Jiaojia belt adopts the evaluation method of mineral resources combining mining area with deposit, and tries to describe the quantity, location, quality and corresponding quantity of deposit. This work is an evaluation work based on a large number of detailed data, and the evaluation method is based on the analysis of geological conditions, which is a comprehensive study of metallogenic controlling factors. It mainly involves five prediction and evaluation methods: geological analogy method, trend extrapolation method, geological geometry method, geological geophysics method and geological geochemistry method.

Geological analogy method is a method to compare key parameters with some highly explored mining areas and evaluate unknown areas. The main purpose of this project is to study the law of mineralization enrichment, establish deposit model and regional metallogenic model, and predict the location and scale of deposits in unknown areas by comparing the relationship between deep ore and shallow ore and the distribution and output characteristics of deep ore.

Trend extrapolation method is the earliest and more mature method in metallogenic prediction. Based on the known characteristics of ore deposits (bodies), according to the natural change trend of related characteristics of ore deposits (bodies), the related characteristics in adjacent unknown sections are extrapolated from known sections. This method is simple, intuitive and effective, and can be widely used in the deep and peripheral metallogenic prediction of mining areas. This book uses the trend extrapolation method to extrapolate the extension depth and scale of the deep ore body according to the change of the external characteristics of the ore body, extrapolate the parameters such as the grade and weight of the deep ore body according to the change of the internal characteristics of the ore body, and extrapolate the deep pinch-out reappearance ore body according to the metallogenic law.

Geometric method is to estimate and predict the mineral resources by geometric method, that is, to describe the ore body with complex shape as a simple geometry, and to transform the complex mineralization state into a homogeneous state within the influence range, so as to achieve the purpose of quickly and roughly estimating its volume and resources. In this work, the block method is used to estimate and predict the amount of resources.

Geological-geophysical method is based on geological exploration research, through studying geophysical fields or some physical phenomena, to infer and determine the physical characteristics of the predicted object, and then infer the geological attributes of the predicted object. Based on the geophysical model established by CSAMT and SIP methods and the predicted ore body position, this project infers the distribution of mineral deposits in unknown areas.

Geological-geochemical method is based on geological exploration, with geochemical dispersion halo as the main research object. By investigating the distribution, dispersion and enrichment of related elements in the crust, combined with geological analysis, it achieves the purpose of predicting mineral deposits (stereology). According to the structural geochemical halo in the well, this project analyzes and judges the position of ore bodies and predicts the distribution of deep ore bodies.