The Northeast of Thailand is located in a tropical climate and dominated by sandy soil types. Owing to the high rainfall (1,162 mm/annum) and low clay content, the soil has low inherent fertility (e.g. pH < 5.5) and soil re-activity (i.e. CEC < 10 cmol(+)kg-1). In order to improve productivity of the main agricultural land uses (e.g. rice), addition of amendments and fertilisers are required to increase pH and exchangeable cations, respectively. Moreover, irrigation and associated infrastructure (e.g. cannals) have been introduced to supplement the variable rainfall. However, the canals are leaking through the low reactive soil, causing recharge water to interact with an underlying sequence of rock salts. The result has been extensive secondary salinization. To better understand the nature and extent of the sandier textured soil, as well as how to manage their infertility and improve water retention in the canals, detailed soil information is required. However, the traditional survey techniques and laboratory analysis are cost and time consuming. In order to develop information to enable soil improvement in terms of adding lime to increase pH and identify leakage areas, more technologically advanced method of digital soil mapping can be considered. Specifically, use easier to measure and acquire digital data to couple this to a limited number of measured soil chemical and physical properties via the use of predictive spatially mathematical models. This thesis therefore focuses on developing digital soil maps (DSM) to create base line information of soil properties including; clay content (%), cation exchange capacity (CEC), and soil salinity. In Chapter 1 the natural resources and land uses of northeast Thailand are described with the need to improve soil condition described. Chapter 2 introduced previous literature on the theory and application of mathematical models, to relate various soil physical and chemical properties to digital data including proximal and remote sensed. In Chapter 3, DSM of clay and CEC are developed at the field scale by inverting EM38 ECa and using a quasi-3D inversion algorithm. The approach is compared with a more commonly used approach of considering DSM of each layer independently. In Chapter 4, a similar approach is used to map soil salinity (ECe) in 3-D, but across multiple fields with a comparison of different EM instruments simulated. In Chapter 5, soil salinity is mapped adjacent to a canal to identity leakage by various ECa data (i.e. EM38, Dualem-421S and EM34) and using quasi-2D inversion modelling. The results confirmed DSM approach can be effectively used to map the spatial detailed soil properties even if in the field or farm scales. Practical guidelines for fulfilment mapping quality over those various morphological properties (e.g. sandy, salinity and acidity profiles) were informatively discussed in addressing and promising the future research. Specifically, for the combination use of available proximally and remotely sensed data with mathematical models including linear regression and machine learning algorithm. Overall, the accurately and economically in producing soil base line information are utmost required to this research.