**4. Discussion**

The two maps shown in Figure 2 are both informative, but a comparison between the kind of information that they provide represents a valuable outcome. The map based on the satellite-collected data provides information about a larger area, and virtually it can provide information about any portion of ground on the planet. However, there are some issues related to this source. First, the satellites cannot always provide pictures of a specific portion of the ground, and therefore a field cannot be continuously monitored through satellite images and satellite-collected data. Second, the resolution of the satellite images is quite low, meaning that a value of CWSI can be computed for each portion of dimension 30 × 30 m. On the other hand, on-field sensors can hypothetically monitor the field 24/7, and the resolution for CWSI maps based on the on-field collected data is quite high, with portions of dimension 1 × 10 m. Such dimensions can be further reduced if the waypoints are placed closer to each other. Therefore, on-field data collection for CWSI maps computation represent a much more reliable source of information on the water stress and health of the plants. It is worth noting that large fields on a homogenous terrain are unlikely to present considerable differences in CWSI values between neighboring areas, but small fields and fields with non-homogenous terrain require a much finer resolution to properly consider the local variation of CWSI in the managemen<sup>t</sup> of the resources and of the field.

The sensitivity analysis by total effects (Figure 3) points out that the most influential parameters are the air temperature (Ta), the UV radiation (UV) and the canopy temperature (Tc). It clearly shows that the accuracy of the corresponding sensors needs particular attention to improve their reliability and lower their uncertainty of CWSI measures.
