Darb Dabney's 3d GIS – OpenSim

This edition of the county-wide terrain model has been upgraded several ways in terms of input data sets, extent on land, and integration of bathymetric soundings to eliminate coastline clipping.

The LiDAR data set from NCALM / GeoEarthScope has been reprocessed to use inidividual ground-classified returns rather than the 50cm gridded surface; this was made possible by NOAA VDatum, the Java application that has industrial-strength powers to position XYZ files.

Also, NOAA Airborne LiDAR for Assessment of Coastal Erosion (ALACE) data, which are first-return only from 1998–2002, were included where available along the Pacific ocean coastline and edges of some estuaries.

An Airborne 1 LiDAR data set was graciously provided by Marin Municipal Water District where it was available in the lower Lagunitas Creek drainage, and this was used where available and not overlapping with the higher-density NCALM data at Point Reyes Station.

More of the FEMA LiDAR data from Dewberry was included along the eastern, urban floodplain areas. In this terrain model, all FEMA points below 25 meters elevation NAVD88, OR points landing on areas with slope less than or equal to 11 percent, were included. In the 2010.01 terrain, only points below 50 feet NAVD88 were used.

NOAA depth surveys were included with very little filtering, for all nearby soundings since 1931—knowing that tidal bars are dynamic, but including all data as a starting point.

The California Seafloor Mapping Project’s phenomenal multibeam sonar work was incorporated from 2-meter grids offshore to California’s 3-nautical-mile limit, and 1-meter grids within west San Francisco Bay.

The earlier photogrammetric breaklines from VARGIS/Infotech were classified into ridge and road types, and only the road types were retained as hard breakline constraints. Ridge lines and water lines were retained as soft breakline constraints. This has mitigated some of the effects of ridge lines artifacts that derive from inconsistencies between the VARGIS photogrammetry work breaklines and contours, and between VARGIS breaklines or contours and overlapping LiDAR data sets.

Gridding the terrain dataset into a dsm, once again we are flirting with the limits of ArcGIS stability. The processing workstation is imaged with Windows Server 2003 to remove limits on output file size, and to permit reliable killing of wayward processes. (At this time, we’re not certain whether this constitutes a “top kill”, or a “static kill”, but in any case the process ends up terminated ;^)

The 2010.07 edition is a topographic-bathymetric surface model, and its prime use case is the generation of accurate synthetic drainage networks. Together, these features motivate a larger modeled extent than the 2010.01 edition. Along the northern area, the extent was grown to include all watersheds that touch or drain Marin County areas, with some clipping of the easternmost portions of the Petaluma River watershed. Offshore, the legal extent of Marin County guided the clipping area in San Pablo Bay; data extent from CSMP guided eastern limits at Richmond Channel, San Francisco’s North Beach and Presidio shorelines, and westerly from Seal Rock to the 3-nautical-mile limit, then northerly to include all of Bodega Head as imaged by NCALM data.

The extra area has made it infeasible to generate a single 40-cm grid as before, but we have cut the area into quadrants and have generated four tiles as 45-cm gridding. More news soon.

Trackback URI | Comments RSS