Geochemical investigations of till is a widely used method in metal exploration as the till commonly inherits the geochemical signature (including the metal contents) of the parent bedrock. In this investigation, over 2000 till samples were collected in the Sarvlaxviken area, southern Finland, where several polymetallic (Cu, Zn, Pb, As, Sn, W and In) veins recently have been discovered in Proterozoic crust along the border between Late Svecofennian granites and the Wiborg Batholith. The bedrock is commonly covered by compact and poorly sorted basal till, formed during the Late Weichselian glaciation event. Several glacial-transported boulders, with high contents of Cu, Zn, As, Sn, Mo and Bi and derived from the local bedrock, have also been discovered on top of the till and provide evidence for concealed mineralisation in the local bedrock under the till cover. The frequent distribution of till in the Sarvlaxviken area provides excellent conditions for the search of such hidden mineralisation by means of systematic till sampling, even if large farm field areas, composed of clay-rich sediments, and seawater-covered areas (Sarvlaxviken bay), had to be avoided in the sampling program. The till samples were collected during university courses and training programs led by the authors and were analysed in a cost-efficient and certified laboratory. Obtained geochemical data were statistically processed by using K-means clustering algorithms which can be used to treat large sets of geochemical data. The results provided anomalies that mainly occur in till with a thickness of <1 m and are considered to be derived from a local bedrock source. The discovered anomalies provide strong evidence for numerous undiscovered veins beneath the till cover.

耕地地球化学调查是金属勘探中广泛使用的方法，因为耕地通常继承母基岩的地球化学特征（包括金属含量）。在这次调查中，在芬兰南部的 Sarvlaxviken 地区收集了 2000 多个样品，最近在 Svecofennian花岗岩和 Wiborg 基石。基岩通常被致密且分类不良的基岩覆盖，形成于晚魏克塞尔冰川事件。几个冰川输送的巨石，含有高含量的铜、锌、砷、锡、钼和铋，来自当地的基岩，在耕地顶部也发现了这些，并为耕地盖下当地基岩中的隐蔽矿化提供了证据。即使由富含粘土的沉积物组成的大农田区和海水覆盖区（Sarvlaxviken 湾），在 Sarvlaxviken 地区频繁分布的耕作为通过系统的耕作采样寻找这种隐藏的矿化提供了良好的条件，在采样程序中必须避免。在作者领导的大学课程和培训计划中收集了样品，并在具有成本效益和认证的实验室中进行了分析。使用K-means聚类算法对获得的地球化学数据进行统计处理，该算法可用于处理大量地球化学数据。结果提供的异常主要发生在厚度<1 m 的泥土中，被认为是源自当地的基岩源。发现的异常现象为耕种盖下许多未被发现的矿脉提供了强有力的证据。