The paper aims at improving a model predicting daily soil temperatures depending on depth and time. The thermal behavior of the ground (near the surface) as a function of depth and time is difficult to simulate from one point since there are many parameters such as short term weather variations, seasonal variations, moisture content of soil, and thermal conductivity of soil etc. affecting on the temperature of ground. The main drawback of this manuscript is that it claims that the improved model will provide the researchers with easily accessible predictions of daily soil temperature variations, which were modeled from daily fluctuations in air temperatures using a sinusoidal function of time and depth. Transient heat flow principles were used with assumptions of one dimensional heat flow, homogeneous soil, and constant thermal diffusivity. Measured and predicted soil temperatures at depths 5 cm, 10 cm, 20 cm and 300 cm were compared with experimental field results to validate the accuracy of the current model. For an annual cycle; at depth 5 cm, 10 cm, 20 cm, and 300 cm the average maximum percentage of errors were 10.78%, 10%, 10.26%, and 14.95%, respectively. Soil temperature measurements at 3 m depth were made on the earth to air heat exchanger system (EAHE) installed in the Solar Energy Institute in Ege University, Bornova, Izmir. Daily average soil temperatures at depths 5 cm, 10 cm, and 20 cm were taken from Izmir State Meteorological Station. Finally, we analyzed solar fluctuations on soil temperature as a function of depth from 5 cm to 300 cm, and time, gave soil temperature as a function of time up to 1 year (8760 h) for the following depths z = 50 cm, 100 cm, 300 cm, 500 cm, and 1000 cm. (C) 2013 Elsevier Ltd. All rights reserved.