Abstract: Abstract: Infrared thermography is a promising technology for crop water status assessment. Effective information can be acquired for the timely formulation of regulated deficit irrigation strategies. However, it is very necessary to optimize the assessment under field conditions, especially under variable environmental conditions (daily and seasonal). Besides, simplicity and robustness are the basis of thermography applicability in the field. In this study, the optimal thermal indicator and the best acquisition were proposed for the monitoring daily period using thermal infrared imaging. The trial was carried out in the solar greenhouse of the National field scientific observation and research station on efficient water use of oasis agriculture in Wuwei from May to September 2021. The tomato plants (Fenxi 5) were selected as the research object. Two irrigation treatments were set: W1-full irrigation (control); W2-deficit irrigation (50% of the control). Firstly, three groups of dry and wet reference planes were selected to calculate the thermal indicators, including the red fabric, green fabric, and artificial spray medium. Subsequently, the performance was evaluated on the four common thermal indicators (Crop Water Stress Index (CWSI), Relative Stomatal Conductance Index (IG), leaf temperature (Tleaf), and the difference between Tleaf and surrounding air (ΔTleaf-air)) in the plant water status diagnosis. Finally, the optimal daily period of thermal imagery acquisition was determined for the tomato plants. The results showed that there were significant correlations of the normalized indices (CWSI and IG) with the plant physiological indicators, such as stomatal conductance (gs), photosynthetic rate (An), and leaf water potential. The leaf temperature Tleaf was also used in a stable planting environment to determine whether the plant was dehydrated or not. There were no significant correlations between the ΔTleaf-air and the physiological indicators of tomato plants under two water treatments. The correlations between the CWSI and IG obtained by the red fabric as the reference plane with the gs, An, and leaf water potential were the most significant among the three groups of reference planes, where the determination coefficients were 0.687, 0.698, 0.669 and 0.707, 0.661, 0.663 respectively. By contrast, the thermal indicators obtained by the green fabric as the reference plane showed a weaker correlation with the gs, An, and leaf water potential, where the determination coefficients were 0.631, 0.655, 0.615, and 0.652, 0.634, 0.638 respectively. The CWSI and IG obtained by the spraying medium artificially were achieved in the lowest determination coefficients with the leaf physiological indicators, which were 0.628, 0.643, 0.609, 0.631, 0.624, and 0.586, respectively. Among the three groups of reference plane acquisition, Twet and Tdry obtained by the red fabric were least affected by the reflection of ambient light. There was no permanent damage to the physiological characteristics of tomato leaves, thereby much easier to distinguish and extract from the thermal imagery. Therefore, the red fabric achieved a great performance to select the reference plane. In addition, the CWSI and IG were most significantly correlated with gs, An and leaf water potential during 12:00-14:00 under both irrigation treatments. Different mathematical functions were obtained to estimate the leaf gas exchange using the best-performing thermal indicators. Therefore, the water status of the plant was effectively determined using thermal infrared imaging.