به‌كارگيري روش آناليز كريجينگ فاكتوري در فيلتر كردن ساختارهاي داده‌هاي گراني‌سنجي منطقه معدني كروميت فارياب

Using the factorial Kriging analysis to filter spatial structures in gravity data acquired from Faryab chromite mine area

با توجه به اين واقعيت كه توده هاي كروميت وزن مخصوص زيادي دارند، در اكتشافات معدني معمول‌ترين روش ژيوفيزيكي كه براي اكتشاف ذخاير كروميت پيشنهاد مي شود، روش گراني سنجي است. در نقشه هاي بي‌هنجاري‌‌ بوگه حاصل از برداشت‌هاي گراني سنجي گرچه تاثير بسياري از عوامل فيلتر و حذف مي‌شود، ليكن بازهم مقادير بازماند به‌دست آمده نتيجه برهم‌نهي چند مولفه متفاوت از حيث ابعاد تاثير و درجه اهميت هستند. براي پردازش و تفسير نقشه هاي بي‌هنجاري‌‌ گراني اصولاً از روش هاي متعددي استفاده مي‌شود. نقطه ضعف فيلترهاي ژيوفيزيكي معمول در نظرنگرفتن ساختار فضايي داده ها حين تفكيك آنها است. ساختار فضايي در واقع بيانگر ميزان همبستگي داده ها نسبت به فاصله بينشان است. آناليز كريجينگ فاكتوري (FKA) اساساً يك نوع فيلترينگ زمين‌آماري است كه قادر است با توجه به تشخيص ساختارهاي فضايي از روي واريوگرام تودرتوي داده ها و تفكيك ساختارها، به تجزيه متغير اوليه به مولفه هاي اصلي آن و يا در واقع عوامل اصلي تغييرپذيري بپردازد. در اين تحقيق، داده هاي گراني منطقه معدني كروميت فارياب با استفاده از روش FKA پردازش و تفسير شده است. براساس بررسي‌هاي ساختاري سه ساختار متناسب با فاكتورهاي ناحيه‌اي، محلي و نوفه‌‌ تشخيص داده و دو محدوده براي ادامه عمليات اكتشافي براساس مولفه ناحيه اي فيلتر شده پيشنهاد شده است. نتايج حاصل از اين روش با نتايج حاصل از فيلترهاي ژيوفيزيكي مقايسه شده است. اين تحقيق به‌خوبي قابليت كاربرد اين روش را در فيلتر كردن داده‌‌هاي گراني‌سنجي نشان مي دهد.

Considering this fact that chromite masses possess high density, the gravity method is the most common geophysical method suggested for prospecting of chromite deposits. Usually, the result of superposition of several factors is observed in the acquired datum, which includes different spatial scales. The observed potential field could be assumed as the sum of the regional field, the residual field, and noise. Despite filtering out several factors to obtain a bouguer anomaly map from gravity survey data, the obtained values are still the result of superposition of several components; these components are different from the view points of scale and importance. Definition and recognition of these components are essential in interpretation of geophysical surveys. There are various methods for processing and interpretation of bouguer gravity anomaly maps. These methods, e.g. potential field filters, are mostly based on mathematical analyses using trial and error technique. There are many different methods concerned with separation of the regional and residual components from the gravity map. Upward continuation technique is frequently used to identify regional anomalies and gravity variations of deeper recourses. The upward continuation is a general filter in processing geophysical data that can remove or considerably lessen the contribution of high-frequency, near-surface, shallow causative bodies from the gravity field, resulting in a smooth field reflecting the deeper causative bodies and/or density structures. This method is applied to separate a regional anomaly from the observed gravity. This filter is a low pass filter since the residual component, which is concerned with local anomalies, can be assumed as high frequency part of the signal. The main weakness of usual potential field filters comes from the fact that they cannot take into account spatial structure of components while filtering them. Spatial structure of a variable is an indicator of the amount of data correlation with respect to the distances between the data. Factorial Kriging analysis (FKA) is principally a geostatistical filtering method that includes classic factorial analysis and geostatistics. The FKA method consists of three basic steps: variogram, factorial analysis and Kriging/co-Kriging. This method • computes of the experimental variograms to choose the number of spatial scales to be considered and fit by theoretical models, (generally linear model of regionalization/coregionalization), • applies the decomposition method on variance-covariance/variogram matrix of spatial components (generally principle component analysis/spectral decomposition), • estimats the regionalized factors in order to determine the relative contribution of each factor for the estimation of a particular location and mapping. Factorial Kriging decomposes the raw variable into as many components as the identified structures in the variogram. The basic step in FKA is experimental variogram calculation and fitting a valid model to this variogram. If the variogram is nested, it can be represented as a combination of several individual components variograms. The FKA method includes two types of univariate and multivariate. In the case of a geophysical variable, the univariate type is applied. Therefore, the variogram in this case can be written as a linear combination of its components. In this research, the gravity data, acquired from Faryab chromite mine area, are processed and interpreted using the FKA method. Based on this study, three components, which may represent regional, local, and noise components are defined and filtered based on spatial structure study. Moreover, two locations are proposed for further detailed exploration considering the extracted local component map. Also, the gravity data are processed using potential field filters. In this regard, different heights are considered in the upward continuation filter method applying on the gravity data, and then, the results are shown in the relevant maps. Low value gravity anomalies can be interpreted as the geological structures having low density or special geometric shapes such as a geo-anticline. High value gravity anomalies can be considered as densie masses like chromite lenses. Finally, in this research work, the obtained results from applying the FKA method on the gravity data are compared with the potential field filtering results using the upward continuation filter method. The basic difference between the upward continuation and the FKA methods is that the latter method takes into account the spatial structure of the data while the former does not. This study clearly indicates the capability of the FKA method in filtering gravity data.