The conventional methods used for determination of soil moisture and electrical conductivity are tedious and laborious. It leads to the imperative need to use soil moisture and electrical conductivity sensor and logging to collect the real time data set. These devices detect the change in saturation and salinity levels of the soil. This data has huge application in the precision agriculture, optimised irrigation, soil moisture monitoring and fertilizer application etc. However, the optimal number of sensors and the associated error curtailment is of great significance but cumbersome. Therefore, this study proposes a benchmarking approach to identify the optimum number of sensors required for field scale operations based on feedback from sensor performance under varying range of working conditions such as saturation percentage, salinity and temperature. The experiments were conducted in controlled temperature conditions varying from 2 to 45˚C. The sensor arrays from minimum of three to nine were grouped to collect moisture, salinity and temperature data and associated error. Overlaying the full-scale error band and 95% confidence interval produced by the sensitivity analysis used in determining the outliers. Analysing the sensitivity plots for various sensor combinations suggested seven sensors as the optimum number to minimise the error. Further, these sensors were deployed in gridded heterogenous medium tank for continuous datalogging to study the variation in salinity.