Fiber‐optic distributed temperature sensing (DTS) has been widely used since the end of the 20th century, with various industrial, Earth sciences, and research applications. To obtain precise thermal measurements, it is important to extend the currently available DTS calibration methods, considering that environmental and deployment factors can strongly impact these measurements. In this work, a laboratory experiment was performed to assess a currently available duplexed single‐ended DTS calibration algorithm and to extend it in case no temperature information is available at the end of the cables, which is extremely important in geothermal applications. The extended calibration algorithms were tested in different boreholes located in the Atacama Desert and in the Central Andes Mountains to estimate the geothermal gradient in these regions. The best algorithm found achieved a root mean square error of 0.31 ± 0.07 °C at the far end of a ~1.1‐km cable, which is much smaller than that obtained using the manufacturer algorithm (2.17 ± 0.35 °C). Moreover, temperature differences between single‐ and double‐ended measurements were less than 0.3 °C at the far end of the cable, which results in differences of ~0.5 °C km−1 when determining the geothermal gradient. This improvement in the geothermal gradient is relevant, as it can reduce the drilling depth by at least 700 m in the study area. Future work should investigate new extensions of the algorithms for other DTS configurations and determining the flow rate of the Central Andes Mountains artesian well using the geothermal profile provided by the DTS measurements and the available data of the borehole.