The authors present the design and instrumentation of ECOSENSE forest, a large-scale environmental monitoring infrastructure that aims at studying forest ecosystem dynamics under climate change. The monitoring system incorporates more than 1100 sensors, generating many data to be processed. Such data may be used for archiving, thus enabling the forest state measurements evolution through time. The main contribution is the implementation of a practical framework for using distributed environmental sensor networks that consists of physical structures, dedicated power supply, communication networks and a data management system. This ECOSENSE forest shows the possibility of monitoring large-scale environments with reliability and resilience.

General comments

The authors are first thanked for their contribution. The manuscript is well written, clear, and comprehensive. The instrumentation presented is particularly impressive, both in terms of the number of deployed sensors and the overall system architecture.

While the main contribution of the article is to present the ECOSENSE infrastructure as a blueprint for similar deployments, the conclusion would benefit from further discussion of the actual use of the collected data. In particular, providing examples of scientific analyses, preliminary results, or research questions already addressed (e.g. part 4) using the dataset would strengthen the paper and better illustrate the value of the proposed infrastructure. 

Furthermore, I have minor remarks and questions:

1. Data heterogeneity management

The monitoring system integrates a large number of sensors measuring different variables, likely with different acquisition frequencies, units, and possibly data dimensionality. The manuscript would benefit from a clearer explanation of how this data heterogeneity is handled. In particular:

- How are different acquisition frequencies managed or synchronized?

- How are measurement units standardized across sensors?

- Are there procedures for harmonizing data structures before storage or distribution?

2. Data format specification

The manuscript refers to a "standardised format" (L.132) and mentions file access (L.380), but the actual data format used for storage and distribution is not clearly specified, if I am not mistaken. Providing details about the file format (e.g., HDF5, CSV, Zarr, etc.) and the general structure of the datasets would improve reproducibility and usability.

3. Integration of external data sources

External data sources are briefly mentioned, such as LiDAR scans (L.281). Is there other external data sources integrated into the datasets ? If so:

- How are these external datasets incorporated?

- How are they synchronized with in-situ measurements?

- Are they stored within the same data structure?

4. Security considerations

Section 2.5 discusses security measures against potential network attacks. 

- Is there any protections against local access threats, from Wi-Fi access points, or LoRaWAN gateways (L.245)?

5. Architecture terminology (Figures 4 and 6)

Figure 4 provides a useful overview of the system architecture, and the consistent color and naming scheme with Figure 6 helps readability. 

However, some terminology may be imprecise, it is a difficult task to formalize accurately the overall process. 

- The term "2. Detection" may not be fully appropriate. In many contexts, detection refers to extracting information from raw or processed data, whereas this stage appears to correspond to data measurement. The term "Measurement" may therefore be more accurate.

- Some labels are not immediately clear. For instance, examples "Selection, Formats" under "7. Processing" could be clarified. At least those single words are not obvious.

- The association of "Broadcast" with VPN is somewhat confusing (see also L.376). A VPN generally provides a secure point-to-point communication channel rather than a broadcast mechanism. Conceptually, this functionality might fit better under "10. Access."

Specific comments

- Figure 3 should be placed after 2.5 because the network infrastructure (right part) is presented in 2.5 and illustrated in Figure 3.

- Figure 4 "8. Broadcast" is not readable (white font color on light gray background)

- L.373: "NodeRed" instead of "NoteRed" (I think)

- L.234: "prevent" instead of "protect" would be more accurate.

- L.396: "ECOSENSE" instead of "EOCSENSE"

This paper addresses the design and the implementation of large scale instrumentation system called ECOSENSE forest. This large-scale environmental monitoring infrastructure aims at studying forest ecosystem dynamics under climate change. The monitoring system incorporates more than 1100 sensors, generating many data to be processed. Such data may be used for post analysis and also for new modeling up to virtual reality developments.

The main contribution is the implementation of a practical framework for using distributed environmental sensor networks that consists of physical structures, dedicated power supply, communication networks and a data management system.

General comments

The manuscript is well written, clear, and comprehensive. The instrumentation presented is diverse, from on the shelf sensors up to in-house developed sensors. The scale of the development in a natural environment is impressive.

Anyway, though the ECOSENSE blueprint is clearly described, some aspects require further explanations.

Detailed remarks and questions:

Large scale instrumentation system are implemented and used in different industrial domains or for transport infrastructures monitoring. Their deployment are still gathered with a maintenance strategy. Could you precise the ECOSENSE strategy implemented for ECOSENSE sensors and the whole system maintenance ?

In field sensors are ageing and could drift with time. Could you precise how you manage for instance periodic calibration or autonomous in-situ solutions to control sensor deviation? Due to the large amount of sensors deployed, it could be interesting to make this description for a macro, a meso and a micro scale for a typical sensor.

How metadata and data management in the ECOSENSE systems takes into account or was influenced by OGC standards, HDF5 group recommendations, ZARR, etc.? Could you precise these aspects in your paper ?

For the reading of the paper, o,e will appreciate to have a separate list of your in-house sensor collection . Could you add such a list and make a schematic view of their location into the ECOSENSE system ?

Could you give an overview of the volumetric amount of data daily or monthly produced (including additional periodic field campaigns)?

Finally, as mentioned in the paper such instrumentation system pave the way to the development of new models up to virtual reality developments. Could you precise geolocalisation and time requirements accuracies required for such research works?