Spatial Interpolation of Soil Depth Using Different Methods and Number of Sampling Points at KOG-MA Watershed, Chiang Mai Province

Pradtana Chumtong; Chatchai Tantasirin

Authors

Pradtana Chumtong Departments of Conservation, Faculty of Forestry, Kasetsart University, Bangkok, Thailand
Chatchai Tantasirin Departments of Conservation, Faculty of Forestry, Kasetsart University, Bangkok, Thailand

Keywords:

Spatial Interpolation, Soil Depth, Kog Ma Watershed

Abstract

Distributed soil depth information is essential data for applying hydrological and landslide models. Data from direct measurement can nevertheless be obtained merely at sampling locations. Spatial interpolation therefore becomes an important technique for generating distributed soil depth information from measured point data. This study systematically measured 103 points to a depth at topsoil and bedrock within a 0.65-km2 forested watershed using a knocking cone penetrometer. Different spatial interpolation methods, including Inverse Distance Weighted (IDW), Kriging, Cokriging (COK), Local Polynomial (LP), and Global Polynomial (GP), and varying number of sampling points, i.e., 86, 46, and 22 points, were employed to generate distributed soil depth information. Comparison of spatial distribution and spatial interpolation error (SIE) obtained using different methods and number of sampling points was carried out based on the correlation coefficient (r), and root mean square error (RMSE). Analysis of variance was conducted to verify the differences of mean square error (MSE). The results revealed that the depths of topsoil ranged between 0.37-2.59 m, with an average ( $\bar{x}$ ) and standard deviation (SD) of 0.94 and 0.39 m, respectively, while the depths to bedrock were 1.13-14.60 m, with x of 5.25 and SD of 3.34 m. These results indicated a higher variance of the depths to bedrock than those to the topsoil. Comparison of different spatial interpolation methods showed that the distribution of soil depths of both layers as obatianed from the IDW method was highly correlated (r > 0.9) to the COK method. SIE of the depths to topsoil as obtained using the IDW method was statistically the lowest (RMSE is 0.1561) only when compared with that obtained using the COK(slp) method. SIE of the depths to bedrock were similar (RMSE ranges from 1.8187 to 1.9170 m) and not statistically different. A comparison using the different number of sampling points of both soil layers illustrated that the correlation among spatial distributions wa not high. The decreasing number of sampling points of topsoil from 86 to 22 points increased SIE only when the IDW and Kriging methods were employed. Using 86 points of depths to bedrock statistically gave the lowest SIE only when the Kriging, COK (up), and COK (slp+up) methods were employed.

References

Penizek, V. and Boruvka, L., 2006, “Soil Depth Prediction Supported by Primary Terrain Attributes a Comparison of Methods,” Plant Soil Environ, 52 (9), pp. 424–430.

Chan, H.C., Chang, C.H., Chen, P.A. and Lee, J.T., 2019, “Using Multinomial Logistic Regression for Prediction of Soil Depth in an Area of Complex Topography in Taiwan,” Catena, 176, pp. 419-429.

Yaotanee, K., 2006, Comparison of Spatial Interpolation Methods Using GIS for Estimating Bulk Density and Saturated Hydraulic Conductivity at Kog-Ma Watershed, Chiang Mai Province, Master of Science Thesis, Watershed and Environment Program, The Graduate School, Kasetsart University. 135 p. (In Thai)

Chankaew, K., Nganpongsai, C., Ruangparnich, N., Sukcharoen, S. and Thangtham, N., 1969, KOG-MA Wathershed Research vol.1: Mountain Watershed Management Research at Doi Suthep-Pui National Park, Mueang District, Chiang Mai Province, Department of Conservation, Faculty of Forestry, Kasetsart University, Bangkok, 234 p. (In Thai)

Konagai, K., Johansson, J., Mayorca P., Uzuoka, R., Yamamoto, T., Miyajima, M., Pulido, N., Sassa, K., Fukuoka, H. and Duran, F., 2004, Las Colinas landslide: Rapid and long-Traveling Soil Flow by the January 13, 2001, El Salvador Earthquake, Geological Society of America, pp. 39-54.

Burrough, P.A. and McDonnell, R.A., 1998, Principles of Geographical Information System, Oxford University Press, Oxford, pp. 132-160.

Chen, H., Fan, L., Wu, W. and Liu, H.B., 2017, “Comparison of Spatial Interpolation Methods for Soil Moisture and its Application for Monitoring Drought,” Environmental Monitoring and Assessment, 189 (10), 13 p.

Ding, Y., Wang, Y. and Miao, Q., 2011, “Research on the Spatial Interpolation Methods of Soil Moisture based on GIS,” The 2011 International Conference on Information Science and Technology, March 26-28, 2011, Nanjing, Jiangsu, China, pp. 709-711.

ESRI, 2018, ArcGIS Pro : How Global Polynomial Interpolation Works [Online], Available: https://pro.arcgis.com/en/pro-app/latest/help/analysis/geostatistical-analyst/how-global-polynomial-interpolation-works.htm. [20 December 2022]

Frolla, F.D., Zilio, J.P. and Kruger, H., 2015, “Spatial Variability of Soil Depth. Interpolation Methods for the Southwest of Buenos Aires,” Revista de Investigaciones Agropecuarias, 41 (3), pp. 309-316.

Li, L. and Heap, A.D., 2011, “A Review of Comparative Studies of Spatial Interpolation Methods in Environmental Sciences: Performance and Impact Factors,” Ecological Informatics, 6 (3-4), pp. 228-241. https://doi.org/10.1016/j.ecoinf.2010.12.003