Stomata are accesses of moisture and CO₂ and gateways for evaporation and gas exchange. Estimating leaf stomatal conductance (g_s) is pivotal for the further estimation of transpiration rates as well as the energy and mass balances between the air and plants. Based on the data collected from the Alxa Comprehensive Observation Field of the Community Hydrothermal Balance in the lower reach of the Heihe River, we measured the g_s of Tamarix ramosissima and analyzed diurnal variations under clear weather conditions using a LI-6400 portable photosynthesis system. Meanwhile, by combining micro-meteorological and physiological data, we modeled the g_s of T. ramosissima based on the three most commonly used empirical models. The results showed that: (1) the diurnal variations of the g_s of T. ramosissima first increased and then decreased. With the gradual enhancement of solar radiation in the morning, the temperature and the transpiration rate gradually increased, and g_s increased accordingly, peaking between 10:00 and 12:00 h. On most observational days, g_s fluctuated to a certain extent around 12:00 h, and this was due to the high temperatures that caused the stoma to close. Afterwards, solar radiation weakened, air temperature decreased, the relative humidity in the air increased, the water vapor pressure inside and outside the leaves decreased, and g_s decreased, which led to a decrease in the transpiration rate. (2) We modeled g_s using three commonly used (semi-) empirical models (Jarvis; Ball-Woodrow-Berry, BWB; and Ball-Berry-Leuning, BBL), and we concluded that the Jarvis model always gives the most reliable performance with a modified coefficient of efficiency (E₁), modified index of agreement (d₁), and determination coefficient (R²) at values of 0.775 and 0.891, 0.887 and 0.945, and 0.590 and 0.645 in 2015 and 2016, respectively. (3) The accuracy between the BWB and BBL models was similar, indicating that there was no obvious difference between vapor pressure and relative humidity that influenced g_s.