Objectives of the service
Current forest monitoring by SEMAS in Pará is hampered by frequent cloud cover and the system is not fully automated. As a consequence, illegal activities in Pará’s forests are not always detected in time. SarVision’s radar based SarSentry forest monitoring system is able to detect and quantify deforestation and forest degradation near real time. The system generates forest change maps and anomaly alerts every 12 days in 15m resolution and with a high accuracy. The SarSentry system is demonstrated to SEMAS for different themes in order to prove its complementary value for the state.
Users and their needs
The main user of the SarSentry system is SEMAS in Pará, but other stated departments in the Amazon area are also interested. SEMAS wants to use SarSentry to increase the temporal monitoring resolution, add the detection of forest degradation, reduce current monitoring costs and make the monitoring more efficient and generate additional revenues with help of improved fiscalization (fines for illegal activities) and by offering monitoring support to commercial forest operations.
In addition, SarSentry provides essential input for REDD programs. SarSentry is used by SEMAS for the following applications:
Early Warning System tool for law enforcement.
Selective logging certification tool.
NRT mapping of deforestation, forest degradation and ecosystem rehabilitation.
REDD project implementation, monitoring and reporting tool.
Modelling tool for carbon and ecosystem accounting.
The other Amazon states have indicated to have similar problems with their current monitoring and are also interested in similar applications. Results of the project are shared with other interested states during outreach workshops.
Two out of 3 test site locations have already been acknowledged by the client, the 3d test site for agricultural-impact related monitoring is still to be selected. The first training is expected in April 2022.
Service/ system concept
The first project activities comprise of data preparation and pilot trials per test area. As preparation, Sentinel-1 data is processed for the period 2017-2021, a 2019 baseline is developed or adjusted and after a calibration and validation of the monitoring results with input from SEMAS, the initial system is developed resulting in historical forest change maps for the period 2017-2021. SarSentry software is installed at SEMAS and the system is adjusted, implemented and demonstrated after a user requirement workshop, for each of the three pilot trials. Basic and technical training is provided to SEMAS staff and during intermediate workshops, progress is to be discussed and evaluated. The results are validated by SEMAS and for the last 6 months of the project NRT forest change maps are produced.
Space Added Value
Images from the Sentinel-1 radar (S1) are used as the core data input for the SarSentry system. Next to S1, use is also made of other sensors for e.g. the development of the baseline in case no existing baseline is used, and for calibration and validation purposes (e.g. Sentinel-2 optical imagery, Landsat-8).
The SarSentry monitoring system is very innovative and outperforms systems from other competitors regarding its high accuracy (for deforestation > 95%) and ability to detect and quantify forest degradation. The latter is not detected by any of the current operational wide area monitoring systems in the world.
The project started on 15 November 2021. The kick-off workshop of the project was held on 4 January 2022. From that date onwards S1 radar data for the already selected test sites in Para have been downloaded and processed. Dirk Hoekman of project partner WUR visited the client SEMAS in Belem from 29 November to 3 December 2021 and conducted a User Requirement Workshop. The system will be adjusted and tested per theme in areas that will be selected in collaboration with SEMAS.