The main objective of this project has been to develop an online cloud based service that allows integration and analysis of Earth Observation (EO) and forest inventory data by using a simple user interface. One of the key objectives of this platform is to provide the forest owners and management companies with easy to use analysis tools that help them to make better decisions and to demonstrate that the use of remote sensing data can successfully facilitate the field data collection.
The general objectives and goals the SATForM3D are:
- Provide the forest owners and management companies with easy to use analysis tools that help them to make better decisions.
- Demonstrate that the uses of remote sensing data can successfully facilitate the field data collection and analysis, making this more robust and efficient.
- Integrate the most up to date remote sensing data and field data measurement techniques (TLS, GNSS device) resulting in a more accurate estimation of the forest assets.
- Demonstrate to the users the benefits of the remote sensing for forest inventory planning.
The specific project objectives are the following:
- Assessment of EO sensors (Geoeye, Worldview, CartoSat, Bing Maps, TerraSAR-X and aerial LiDAR) for forestry analysis (see below bullets 2,3 and 5)
- EO analysis for automated recognition of forest canopy and "productive area" extraction
- EO analysis for automated stratification of the forest stands based on the forest canopy characteristics.
- Automated system for creating inventory plans (plot samples)
- Automated forest canopy change detection for clear-fell and windblown areas.
- Assessment of GNSS receptors, in order to improve the precision in the field measurement positions.
- Remotely send and receive field data from the Treemetrics Forest application.
- Automated printable reports generation showing the 'productive area', 'stratification' and 'inventory plots'
- Integrate the field data collection from forestry inventories with a remote database and analytic platform improving the data integrity.
- Integrate the remote sensing data with the TreeMetrics 3D laser scanning measurement system and the ‘Forest WarehouseTM’ valuation and optimisation tool.
Users and their needs
The targeted users of this project are mainly forest owners and management companies. Forest owners group includes state or public companies, private forest companies or individual forest owners. It is common for forest management and consulting companies to be in charge or the forest management, providing the tools necessary for meeting the owner needs. These companies provide several services as forest inventory, management planning, harvest planning, timber sales, etc.
The project stakeholders have been categorised into three major groups:
- Smallforestowners: owners of small forest areas (<50ha) and typically with limited resources.
- Largeforestowners: are private or public owners, usually forestry companies or state agencies with vast forest areas.
- Managementscompanies:range in size from small to large. Management companies can help large groups of small owners organised into grower associations. This form of structure is becoming increasingly popular in regions of the world with many numbers of small fragmented owners.
In addition, state agencies and forestry associations have also been representative users in the project:
- State agencies: Forestry and Environmental agencies; typically responsible for matters relative to forest areas monitoring at national level.
Forestry associations: represent forestry owners and give advice to technicians and forest owners.
During the SATForM 3D project, representative needs were gathered as a result of the deep analysis and feedback from the project partners, including representatives of the different Target Users.
- The users, especially small companies and forest owners with not GIS knowledge, need to have an automated system to analyze EO data for forest canopy identification (productive area) and identify forest zones (stratification).
- The users need to have available EO data less than on year old in order to get a realistic productive area and forest stratification.
- The user need to generate more automated reports including forestry statistics and resources maps for each forest unit, crating standards and reducing time in maps and statistics generation.
- The users need to optimize the data transfer (e.g. DBH, height, stocking and locations) form the field to the office and vice versa following standard formats that does not require travelling to the office or manual data entry to the database.
- The users need to have the data secure, but easily accessible.
The users need to share information with other stakeholders such as customers, providers and other company workers
Service/ system concept
Forest Mapper is accessible through a web-based platform, with a service offer provided in an descriptive and visual manner and which includes tools for:
- Automated recognition of forest areas and "productive area" extraction.
- Automated stratification of the forest stands based on remote sensing.
- Automated system for creating an inventory plan based on remote sensing.
- Automated forest canopy change detection of clear-fell and windblown areas.
- Remotely send and receive field data from the Treemetrics Forest application.
- Automated printable reports generation that meet forestry industry standards.
- Store document associated with the forest or forest operations
- Successfully integrate the remote sensing data with the TreeMetrics 3D laser scanning measurement system.
- Analyze inventory data in order to create forest strata and create reports showing these results.
- Perform forest assets forecasting based on Irish growth models.
Space Added Value
The SatForM 3D demo project was focused on demonstrating the integration of Satellite Earth Observation (EO), Satellite Navigation and ground based remote sensors (TLS) for fully automated forest inventory analysis. On top of this, Forest Mapper architecture maintained the capability of including data from forest surveys based on manual data collection (calliper and hypsometer).
Within the SatForM 3D project different sensors have been tested (Bing Maps, Geoeye, Worldview, Cartosat and aerial LiDAR) -" in order to assess the suitability of each sensor for the different EO analysis included in Forest Mapper (see above). In addition, datasets from TerraSAR-X were studied by the University College Cork (UCC), with not conclusive or useful results for the analysis proposed.
The performance of the different sensors was similar. Bing maps was more detailed than the rest because it has more resolution (0.35m) than the EO sensors (0.5m), however Geoeye and Worldview have been more reliable than Bing detecting vegetation because the Near Infrared band, this is a requirement for the change detection tool.
Cartosat stereo image and LiDAR have performed well and they are more reliable in forest detection due to the height information. However, broadleaf and conifers were misclassified in some cases. In addition, Cartosat needed some corrections using the Digital Terrain Model (DTM) from LiDAR as reference.
Based on these results and the user preferences, Bing is a preferred solution in the immediate future due to the more detailed result and the immediate accessibility to the data source. A decisive factor for this choice is that currently the Bing imagery available for Ireland is only one year old, however this can change in the future because the Bing Map acquisitions is uncertain. Geoeye has been the second choice from the users, in addition, this sensor is required for change detection. Due to the Bing Maps acquisition issues is expected that Geoeye preference would progressively increase. LiDAR demonstrated that results based on height information (DHM) can provide successful results, however the users did not selected this as a preferred option because the current availability and cost.
The use of the GNSS device is a requirement for most foresters in order to record position of the forest field measurement. During the SatForM 3D project, different high precision GNSS device were tested.
Three GNSS devices were tested: (i) Trimble GeoXT with Floodlight, (ii) U-blox and (iii) Asus Nexus 7. The fist two devices are categorized as high precision devices with sub-meter accuracy. The last GNSS receiver is the default received built in the Nexus 7 tablet for comparison purposes.
The results show that GeoXT have a lower Horizontal Dilution of precision (HDOP) than the other two devices in open sky, however under forest canopy U-Blox have a better accuracy in HDOP. According to this U-Blox was preferred under forest canopy, therefore selected as the GNSS device used in this project.
The main benefits that the SATForM 3D project is expected to deliver to the forestry sector are:
- Easy to use: Forest Mapper has a very simple and intuitive user interface that allows the user to navigate and perform analysis without needing previous technical knowledge.
- Accessible anywhere: Forest Mapper can be accessed from the office, conference room or even from the mobile phone. The only requirement is to have internet access.
- Up to date remote sensing datasets: Forest Mapper provides the user with a wide selection of different remote sensing datasets form different providers (Geoeye, airborne LiDAR, Bing Map and possibility of new providers).
- Independent analysis: Forest Mapper provided automated EO analysis, independently of the user experience.
- Reliable and automated transfer of field data: Forest Mapper connection with Treemetrics ForestTM application allows the transfer of field data automatically using the internet network.
- Platform for sharing information: Forest Mapper allows the user to share the project with other Forest Mapper users in a "viewing" or "editing" mode. Therefore, projects can be shared with customers, providers or other workers in the same company.
- Security and Data Ownership: Forest Mapper is a secure system with different user profiles. Each user has restricted access to the remote sensing data, projects and tools that they can access according to their privileges
- Forest valuation: Forest Mapper has the added value of the Forest WarehouseTM link. This is a unique tool for Terrestrial Laser Scanner (TLS) analysis and cutting optimization.
Forest MapperTM is based around an interactive map where the results of the analysis are displayed and where the user can retrieve information and features, as well as editing. This helps the users to identify areas where action is necessary and will encourage the industry to adopt new technologies to better manage the forest resource.
Forest Mapper has a very simple and easy to use user interface that allows the user to navigate and perform analysis without need previous technical knowledge in GIS or statistic. The main features of the system are the following:
- Remote sensing datasets available : Forest Mapper provides the user with a selection of different remote sensing datasets(Bing Maps, Geoeye and LiDAR) and the possibility to acquire similar type of data from other providers in the future.
- Forest EO analysis : Forest Mapper is able to analyse automatically the Earth Observation (EO), described above, and automatically generates forest productive area, forest stratification, forestry inventory plans and harvest area detection.
- Distribution maps : Forest Mapper displays the forestry parameters collected in the field (i.e. DBH, tree size, timber volume, etc) in distribution maps that allow the user an intuitive interpretation of how the forest resources are distributed within the forest.
- Forest inventory - timber survey -: Forest Mapper can be connected with the Treemetrics ForestTM application (www.treemetrics.com). It allows the user to download maps to the field device and upload the field data back to Forest Mapper, once the measurements are collected. If the user is not using this application, printable reports can be generated.
- Forestry analysis: Forest Mapper provides a complete forestry analysis parameters (i.e. mean DBH, average tree size, timber volume, etc) from the forest inventory data. This system use the areas generated by the EO analysis to estimate the total resources in the forest. The data can be display in tables for different forest units and using distribution maps. A thinning analysis option is also available.
- Forest forecast: Forest Mapper will use the Irish yield models to predict the future production of timber based on the current forest status.
- Reports: Forest Mapper generates both standard and customisable reports showing the results of the analysis described above.
- Sharing information: Forest Mapper allows the user to share the project with other Forest Mapper users in a "viewing" or "editing" mode. Therefore, projects can be shared with customers, providers or other workers in the same company.
- Secure: Forest Mapper is a secure system base on Ruby on Rails and Devise, including different user profiles. Each user has restricted access to the remote sensing data, projects and tools that they can access according to their privileges.
Current systems for forestry assessment available to the different users have a big opportunity for improvement. Among the areas for improvement, error in the area estimation is one of the biggest problems for foresters; together with out of date images and unrecorded forest disturbances. Despite the fact that ground surveys take place in order to generate forest field inventories, the size of the forests, the low visibility and the difficult terrain can result in some areas never being surveyed.
Up to now the majority of individual forest owners and forest managers have worked based on out-dated maps (i.e. more than 2 years old) that do not represents the current state of the forest. Additionally some of the maps used by individual owners and companies have very limited information and typically are provided in hard copy form. This is a current problem to forest owners bringing their wood resource to the market as the information needed by a timber purchaser to provide a fair bid is insufficient.
In the last years, global web mapping services, such as Google Maps or Bing Maps have facilitated the access to some type of imagery that the forester may use as guide for forest inventory planning or to locate damaged areas in their forest. However, there still remains a lack of proper analysis in order to provide a real value of the forest productive area. Currently, the capability, the access to and the analysis of Satellite Earth Observation data (EO) is mostly inexistent for small management companies and forest owners.
A key issue of global web mapping services is that they are not targeted to forestry, and therefore they do not always provide up to date imagery and/or the required level of spatial resolution to allow for a satisfactory analysis of the forest areas. These web services only provide regular imagery (3 bands). In order to overcome these limitations the use of other providers of EO data is necessary.
The main advantage of the satellite EO is that these can be acquired relatively easily on demand if necessary, even in remote areas. Current EO imagery available includes typically from 4 to 8 bands (e.g. Geoeye and Worldview), this is an important added value for forestry analysis and vegetation analysis. Otherwise, the main disadvantage is that the resolution is limited to 0.5m for commercial use and acquisitions time required varies form 30 to 90days.
In addition to satellite EO, airborne LiDAR can penetrate through the canopy, this fact enables the possibility of additional analysis, such as the tree height and under canopy analysis, improving the accuracy and reliability of the automated analysis.
Forest Mapper allows the use of different remote sensing sources, including 3 and 4 bands imagery and airborne LiDAR, depending on the availability.
- Most of the Demo users (nearly 100%) are "satisfied" or "completely satisfied" with Forest Mapper and all agreed that the product developed is of "above average" or "excellent" quality.
- Currently the preferred remote sensing dataset by the user is Bing Maps, due to the majority of users already being familiar with it and due to the high spatial resolution provided in Ireland. This is influenced by the fact that Bing Maps has acquired new aerial photography for the entirety of Ireland one year before the user Demonstration started. The second preferred sensor is GeoEye, the data from this sensor is more up to date than Bing Maps, but it has a longer time for acquisition and a lower resolution. It is expected that this preferences will change over time, if Bing Maps data in Ireland is not updated regularly.
- All users found an added value to the integration of EO analysis and forest field measurements in one online platform in order to better manage and estimate the timber resources in the forest areas.
- It has been demonstrated that Forest Mapper can save significant time for the management companies regarding forest area delineation, automated forest inventory analysis and the automated generation of standard reports for timber survey results, pre-thinning analysis and timber growth forecast.
- The small owners found an added value in the Irish