S ite-specific farming involves managing each crop production input on a site-specific basis to reduce waste, increase profits, and maintain the quality of the environment (Morgan and Ess, 1997). Use of the global position system (GPS), geographic information systems (GIS), and variable rate applicators (VRA) make it possible to individually treat different spots in the field. In order to vary fertilizer rates, information about the variability of soil properties is needed throughout the field. To date, the most common technique of soil analysis is through grid sampling. In this case only one averaged value typically represents each soil property in an area up to 1 ha. Also, current methods used in many soil laboratories are time-consuming and expensive. Thus, there is an ongoing need to develop automated systems to decrease the cost of soil sampling and improve the accuracy of soil nutrient maps (Sudduth et al., 1997). Soil acidity/alkalinity (evaluated by soil pH) is one of the basic chemical properties. Soil with low pH contains acidic ions at levels that can be toxic for some plants. Also, soil acidity limits the amount of basic ions such as calcium and magnesium, and reduces the availability of nutrients such as nitrogen, phosphorus and potassium, slowing plant growth. In addition, soil with low pH can reduce both the effectiveness of some herbicides and activity of bio- organisms (Mengel, 1997; Thomas, 1996). Low soil pH should be corrected by liming. Various methods have been developed to provide quantitative measurements of soil acidity and limestone needs. Most of them include determination of both soil and buffer pH. Alternative methods use relationships between soil properties such as texture, organic matter content, clay mineralogy, and buffer capacity to predict lime requirement in the field (Hergert et al., 1997). The goal of this research was to design and test an automated soil sampling system to estimate soil pH on-the- go. The proposed method was based on measurement of soil pH at field moisture contents with a flat surface, combination pH electrode. LITERATURE REVIEW SOIL SAMPLING In the past, only a few randomly selected samples were used to estimate the soil pH of the field. Today’s agriculture has been introduced to precision farming techniques, which help farmers consider the variability of pH (or nutrient) levels within a field and manage them accordingly. Soil pH may have significant variation within a field. Some fields may have soil pH ranging from 5 to 8, and the coefficient of variation can be over 10%. (Borgelt et al., 1994; Mulla, 1993). Combining soil cores over an area of 1 ha leads to a loss of information about spatial variability. Many questions still exist about the accuracy of interpolated maps from grid soil sample data. Also, errors can originate from sampling and analytical methods, sources of which can be numerous (Wollenhaupt et al., 1997). To select an appropriate sampling program, it is necessary to consider soil CEC, nutrient levels, crop rotation, etc. In general, current crop recommendations for AN AUTOMATED SAMPLING SYSTEM FOR MEASURING SOIL pH V. I. Adamchuk, M. T. Morgan, D. R. Ess ABSTRACT. Within the scope of precision farming there is a need for improved methods of assessing and managing soil variability. The site-specific management of soil pH is one application that has potential benefits for crop production. However, current grid sampling and mapping techniques to estimate lime requirement may not be adequate. For this reason, an automated soil sampling system for measuring soil pH on-the-go has been created. The system includes a computer-operated soil sampling mechanism mounted in a shank, a global positioning unit, and a pH meter with a flat surface electrode. The system measures soil pH directly on a sample. The automated soil sampling system can determine pH while taking soil samples at a selected depth (0-20 cm) every 8 s. A simple linear regression was used to calibrate the electrode mV output against soil pH obtained via a standard laboratory method. Field testing yielded an r 2 of 0.83 and a standard error of prediction of 0.45 pH. Keywords. Precision agriculture, Soil pH, Soil sensors. Article was submitted for publication in September 1998; reviewed and approved for publication by the Power & Machinery Division of ASAE in May 1999. Presented as ASAE Paper No. 98-3094. Approved as Paper No. 15788 of the Purdue University Agricultural Research Programs. Mention of a trade name, proprietary product, or company name is for presentation clarity and does not imply endorsement by the authors, Purdue University or University of Missouri, or exclusion of other products that may also be suitable. The authors are Viacheslav I. Adamchuk, ASAE Student Member, Graduate Research Assistant, Mark T. Morgan, ASAE Member Engineer, Associate Professor, Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, Ind.; and Daniel R. Ess, ASAE Member Engineer, Department of Biological and Agricultural Engineering, Assistant Professor, University of Missouri, Columbia, Mo. Corresponding author: Mark T. Morgan, Purdue University, Dept. of Agricultural and Biological Engineering, West Lafayette, IN 47907-1146; voice: (765) 494-1180; fax: (765)496-1115; e- mail: mmorgan@ecn.purdue.edu. Transactions of the ASAE © 1999 American Society of Agricultural Engineers 0001-2351 / 99 / 4204-885 885 VOL. 42(4): 885-891