Oahu hydro enforced digital elevation model 2m v1.0

Description

Technical info

Background value: -32767

Max: 1232.3788 m

Min: -2797.7319 m

Universal Transverse Mercator (UTM) NAD83(PA11)

EPSG:: 6634
units: meters (m)

Vertical Reference Frame - LMSL

32bit raster grid of floating point values representing elevations for the island of Oahu. Coverage extends to nearshore and coastal areas around the island.

Note that ground resolution for the digital elevation model (DEM) is approximately 2 m in the planar (UTM) coordinate system.

The Oahu_HydroEnforced_DEM_2m_v1.tif was developed for input into coastal flood models for the island of Oʻahu. 

The DEM incorporate UH SOEST HMRG, NOAA NCEI, NOAA OCM, USGS 3DEP and USACE JALBTCX derived elevation data from LiDAR surveys, prior survey efforts and spot elevations. The layer organization below was used to initially set up the hierarchy of data sources:

DEM construction:
Layers (bottom to top order):
himbsyn.bathy.v19.gmg - UH SOEST HMRG
NCEI_ninth_Topobathy_Hawaii.9428 - NOAA NCEI CUDEM
hawaii_bty_5mGM - UH SOEST HMRG
USACE_2007 - USACE JALBTCX
Oahu_2013USACE1m_DEM_LMSL - USACE JALBTCX
NOAA_Oahu_DEM_2013 - NOAA OCM

Layers were ordered based on confidence and resolution. USACE 2013 data was used for nearshore from -40 meters to approximately 2.5 meter elevation where coverage existed. NOAA 2013 terrain data was used from 2.4 meters and higher terrain were coverage existed. NOAA NCEI CUDEM and USGS 3DEP was used to fill data gaps onshore. UH SOEST HMRG and NOAA NCEI CUDEM was used offshore below 40 meter water depth. Overlap of data layers and stable features was used to help resolve differences between terrain datasets and reprojected bathymetric datasets (WGS84 for CUDEM and HMRG and NOAA Nautical Chart data). Model uncertainty was also considered. Continuous coverage was emphasized in the creation of the final elevation surface. Seams between datasets were smoothed using a weighting function or interpolation if data gaps existed. Large ‘steps’ between datasets can impact some model’s performance.

Waterways and inlets were identified and checked to assure depths reflected existing data sources (2013 USACE data and NOAA nautical charts) where data overlapped. USGS breaklines were used to flatten inland water bodies. Data gaps were filled using a combination of interpolation between identified 'good' data and down-sampling. Artificially high raster values that overlaped water bodies were lowered using digitaized polygons of the water feature and estimated elevation values. Streams, culverts, ditches and canals were assigned with single values or using lines and slope values between elevation considered valid. No fieldwork was performed for this project to verify elevation data or the edits made.

Other data considerations:
- Across the dataset gulches and canals were reviewed for connectivity to the ocean. In many cases, manmade features (removed - interpolation artifacts, and unremoved bridges, structure footprints) and misclassified vegetation were removed. Artificially high areas were lowered using sloped lines and polygons to join drainage infrastructure and remove interpolation artifacts that prevented water from moving downhill.
- building footprints located below 20 meter elevation were reviewed and minimum mean elevations of surrounding terrain were applied as 'foundation' elevations. The minimum mean elevation was defined as the mean of the 'ground' within the 2D footprint space and is used to flatten all building footprints using a single elevation around the island, where they exist. see Notes 1, 2 and 3 below. 

Notes: 
1) During classification of the LiDAR, the building returns, where they existed, were removed.  
2) The methods used to process the 2013 USACE and NOAA LiDAR data into raster DEMs varied. Specifically how data gaps where buildings were removed were treated. USACE raster data includes triangulation artifacts while NOAA use single values for most removed buildings. 
3) Structures were downloaded from the City and County of Honolulu Geodata server on Aug 19, 2022 and features were geometrically fixed/repaired using QGIS - 'fix geometries' tool. Buildings that overlapped floating piers or water bodies were removed from the dataset. Building footprints that were only partially in the focus area (below 20 m elevation) are not included in the analysis. The dataset does not appear to be maintained with accurate building information such as min elevation or building height.

Notes from processing:

1) Interpolation artifacts exist where channelized streams pass under removed highway feature
2) Channelized streams in higher elevations were reviewed and some cases ‘carved’ through roads and misclassified vegetation to provide connectivity. In some cases, the slope and location of these simplified features do not likely reflect the reality and should not be considered accurate. They do enforce hydro-connectivity where information is incomplete.
3) At the airport reef runway service road, channels were carved to represent the connectivity of the lagoon to the ocean based on imagery and visual inspection.
4) Changes and updates to streams and artifacts in the DEM were corrected using a line slope method to replace interpolation artifacts (artificially high elevations) with slope values from upstream and downstream minimum elevations of the channel. 

Created: 09/19/2022
Edited: 02/11/2025
Editor: Matthew Barbee
Readme Version: 4.0

Data release version: 1.0