An improved core sampling technique for soil magnetic susceptibility determination

• An improved core sampling technique for determining magnetic susceptibility in soils was evaluated.
• The technique saved considerable time and labor when analyzing large sample sets.
• Data accuracy of this technique was satisfactory compared with the conventional auger sampling method.
• The technique was for rapid magnetic susceptibility determination in soil related research at large spatial scales.

The efficiency of acquiring soil magnetic susceptibility (MS) data is strongly influenced by time-consuming sampling and treatment procedures of the conventional MS technique (M1). It is necessary to find an optimized technique to save time and labor in soil-related research, especially at large scales with a large number of soil samples. An improved soil sampling technique (M2) based on an improved core sampler kit and a simplified sample treatment procedure was introduced. Sample treatment of M1 is similar to other common soil- and sediment-processing procedures, while M2 involves relatively limited procedures without grinding and sieving. A total of 360 samples from 5 soil layers at 12 sampling sites based on M1 and M2 were collected on a gentle cultivated slope in Northeast China. The tested black soils contain few gravel (> 2 mm) and other non-soil components. Results showed that M2 saves a considerable amount of time and labor for MS determination compared with M1. Assuming one person, 85% of the time processing 10,000 soil samples using M1 can be saved by using M2. Additionally, M2 produces MS data (χlf, χfd%) with satisfactory accuracy compared with M1. Pearson correlation coefficients for χlf and χfd% between M1 and M2 are 0.973 and 0.882 (n = 180) at the 0.01 level of significance, respectively. The mean relative error in χlf and χfd% based on M2 is 8.5% and 8.2%, respectively, when compared with M1. These results indicate that M2 for MS determination is effective for soil related research at large spatial scales where large sample sets are required or representative samples are required in advance for other time-consuming or expensive analyses.