The following information is provided for users to assess the usefulness of the data to their own needs. 1.) Terrain Mapping Northern Ontario Radar Digital Surface Model has been used by MNDM daily in the Terrain Mapping Project for many different tasks in their remote sensing analysis and in characterizing the various landform features and sediment distribution in the project area. The main uses made of the DEM include: identifying landforms and hence material types with the ability to identify and map some very small forms such as the bioherms along the Attawapiskat River (commonly only about 200 X 400 m and range from 2 to 8 m high); certain sediment types have distributions that are elevation dependent such as the distribution of marine and glaciolacustrine origin; it will be extremely useful in determining the regression of the Tyrell Sea which effects the length of time that say carbon sequestering has been going on in the various regions of the Hudson Bay Lowland when shoreline elevation and basal dating of wetland deposits are combined; the Radar DSM has been used daily in the segmentation and classification of Spot imagery tiles (in particular, derivatives such as the soil wetness index, slope, aspect, etc.); and it has commonly been used in the creation of linear symbols that occur on the maps (such as defining river terraces, shoreline features etc.). 2.) Landcover Classification IMA conducted a comparison among the original source data (1 arc second spaceborne C-Band interferometric radar data), Northern Ontario Radar DSM (this product) and the federally derived 1:50,000 DEM available from NRCan in land cover mapping. The comparison revealed that, although noisy, the horizontal precision of the 1 arc second original source data allowed for far more precise mapping of the topographic features associated with major ecological breaks (i.e. upland/lowland). IMA used it to isolate upland and riparian areas to augment land cover mapping. The original source data was noisy, and incomplete with a few scattered areas containing tiles derived with the federal contours. To make use of the original source data, IMA has to deploy certain quick yet effective measures to smooth vegetative noise using a simple 2 pass median filter (applying a median filter with 5 by 5 window twice). However, for NRCan DEM that were clearly contour based, manual editing needs to be employed, which is very time consuming and subjective. On the whole even the original source data was far more detailed than the national DEM provided by NRCan. Upon receiving Northern Ontario Radar DSM (this product), it became apparent that the NRCan DEM tiles originally derived with contour data had been replaced with the new product. Further, the Northern Ontario Radar DSM maintained the integrity of the original source data (1 arc second spaceborne C-Band interferometric radar data) very well, which allowed the inclusion of more vegetative and topographic precision than our original filtered product (produced by using a simple 2 pass median filter on original source data). This detail will be helpful to IMA to move forward with interpretation efforts for land cover accuracy assessment. Additionally, the vegetative detail will allow scientists to better estimate vegetative structure while the horizontal precision will improve our understanding of topographic expression. The entire Ontario's Far North land cover classification beifited from the Radar DSM. Currently it is the only way to conform to the landcover mapping standards set by IMA. According to IMA, the DSM which represents the one consistent elevation product for the entire province was essential input into our land cover mapping model. Without it we would not have been able to generate a land cover that would adequately meet Far North planning requirements. By generating a wetness mask from the DSM we were able to separate upland from lowland vegetation classes, thereby elevating our product from a basic vegetation map to an ecological land classification map that conforms to Ministry standards of ELC Community Class Heirarchy. This year we have been asked to extend our mapping further south to cover the areas of the proposed transportation routes leading to the Ring of Fire area and will once again make good use of the radar DSM. 3.) DSM hydro-conditioning WRIP is working on hydro conditioned DSM based on Northern Ontario Radar DSM (this product). Elevation vector points are extracted from Northern Ontario Radar DSM, and re-construct the hydro-conditioned DEM using river vector lines (hydro network), which is originally provided by NRCan and edited and refined by MNR. The original source data is not suitable for hydrological applications. WRIP is now using the Radar DSM for hydro-conditioning. Overall, the Radar DSM product was found to represent ground topography and geomorphology very well. 4.) Community-based Land Use Planning WRIP staff has worked closely with the Cat Lake / Slate Falls planning team through 2010/11 to illustrate the potential use of the Northern Ontario Radar DSM and hydrology data products being created. This involved reviewing the topography in the planning area using the DSM. The hydro conditioned DSM is used to create watersheds for the planning area. Far North planning staff and First Nation representatives identified that the DSM, the hydrology network and watershed data were fundamental for their planning process. Watershed boundaries were used to identify land use planning boundaries for the official Cat Lake / Slate Falls plan finalized on July 11, 2011. It is anticipated that the hydro conditioned DSM and derivatives created using the Radar DSM will be key elements to all 38 Far North communities as planning efforts continue across Ontarios Far North. In total, 47 tiles of 250K maps and 153 tiles of 100K maps have been made in GeoPDF format using the Radar DSM as background for community based land use planning.
Process step:
The steps to generate the Northern Ontario Radar DSM are as follows: 1. Extract elevation vector points from raster source data. In order to keep full precision of original source data, raster data is converted to vector points before projection. 2. Mosaic source data tiles of vector point into three shape files. 3. Project the vector point shape files from Geographic Coordinate System (GCS) to MNR standard Lambert Conformal Conic (LCC) projection. 4. Interpolate the projected point data into raster grids using ESRI Local Polynomial Interpolation (LPI) algorithm. 5. Control cell origin to align up with current provincial DEM standard 6. Merge the subregional raster grids into one seamless grid. 7. Convert the elevation vertical datum from EGM96 to CGVD28 (this product is avalable in EGM96 and CGVD28 vertical datum).
The source data used in this project for generating Ontario Radar DSM is acquired using spaceborne C-Band interferometric radar, which collected Interferometric Synthetic Aperture Radar (IFSAR) data during an 11 day mission in February 2000 from the Space Shuttle Endeavour. The mission performed a single pass radar interferometry from two antennae mounted 60 m apart with the swath width of 225 km, which cover 80% of earth, and 95% of residential area of the world. The tile of the source data is processed in one degree by one degree extent. The edges of each tile are matched with the edges of adjacent cells to ensure continuity. The source data tile is a uniform matrix of elevation values indexed to specific points on the ground (nominally 30-meter post spacing). The World Geodetic System 1984 (WGS84) is used as horizontal datum, and WGS84 Earth Gravitational Model (EGM 96) geoid is used as the vertical datum to represent mean sea level. The actual spatial resolution is 1 by 1 arc second below 50 degrees north (3601*3601 elevation values per tile) and 1 by 2 arc second north of 50 degree (3601*1801 elevation values per tile). The pixels of the source data represent elevation values of the reflective surface, which could be: 1. Vegetation (e.g. Tree fen); 2. Man-made features (e.g. Abitibi Canyon Generating Station); 3. Bare earth (e.g. folding bed rock). 211 original source data tiles were used to create the Ontario Radar DSM covering the entire province of Ontario. To create a seamless DSM in ESRI ArcGrid data format (in both EGM96 and CGVD28 vertical datum separately).