CICEET Progress Report for the period 3/15/05 Through 9/15/05

Project Title: Salt Marsh Assessment and Restoration Tool (SMART)
Principal Investigator(s): Jeffrey N. Rogers
Additional Investigator(s): Ray Konisky; John Mustard
Project Start Date: 10/1/04

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Tasks to meet objectives

• Subcontracts with Brown and Hyperspectral provider
• Flight and field data planning/acquisition
• Capture of LiDAR and Hyperspectral Imaging
• SWMM model refinement
• Begin SMART extension programming

Progress on Tasks
Project sub teams began working in a greater capacity on the four areas of research (Hyperspectral, LiDAR, SWMM, and SMART). Subcontracts were issued to Brown University and Flight Land Data completing the project contracting. LiDAR data was collected by C-CLEAR, a Canadian research group from Nova Scotia, in mid April at two resolutions (1.5m and 0.3m postings). Bare Earth plots were used to help control sensor datasets. These were 8x8-foot areas cleared of all vegetation and thatch (Figure 1). Early season Hyperspectral imagery was collected in early June by Flight Land Data of North Andover, MA. Extensive ground truth data was collected in support of both LiDAR and Hyperspectral flights including Kinematic GPS (Figure 2). Dr. Mustard and Mr. Fisher began processing of the hyperspectral datasets to create a usable image for the implementation of vegetation extraction techniques. Mr. Rogers and researchers at UNH have begun manipulation of the LiDAR dataset acquired last spring. This includes statistical analysis of the sensor data and the ground truth by vegetation type.

Mr. Quigley and other GeoSyntec staff have continued to work on the SWMM hydrologic model of the Drakes Island Marsh system. They have now refined it to include precipitation and marsh topography (stage storage). We have also updated the model to existing conditions after the culvert was replaced this spring. We have been collecting water level calibration data since the ice broke last spring. Progress in the Software Development Tasks included acquisition of the project software development environment (ESRI ARCGIS, ESRI Spatial Analyst, and Microsoft Visual Basic). The source code modules for an existing estuarine Visual Basic modeling extension was acquired from the NOAA Coastal Services Center and installed on the development computer. A review of source code and technical documentation with the extension author at NOAA CSC helped frame out the technical approach for implementing SMART in ARC GIS. Also a new release of the earlier model application (Marsh MD) was developed to streamline internal model pro

Difficulties
The majority of our difficulties for this reporting period are related to remote sensing acquisitions. We have had continued trouble getting the planes to fly exactly when we want them. This is inherent in this industry but in addition to normal delays our providers have had continued aircraft trouble. Also a miscommunication caused our LiDAR provider to send their sensor away for three-four week period during our scheduled acquisition. Aircraft problems and increased inclement weather has caused us not to have capture our second LiDAR flight as of yet. There is hope that it will be captured tomorrow as now scheduled. We have also had difficulties with our Hyperspectral provider, Flight Land Data (FLD). The data quality, resolution and expected work product were not up to the specifications in our subcontract agreement or industry standards. We have spent a significant amount of time trying to correct and compensate for this but believe we have made it acceptable for our uses. We have terminated our contract and FLD did not fly the site a second time. Fortunately the Army Corps of Engineers flew the site for their own purposes with their Hyperspectral sensor in September and have offered us access to the data for our project.

Project Objectives for Next Reporting Period

Objectives
The objectives for the next reporting period are to develop vegetation extraction methods from the two Hyperspectral dataset, continue to analyze the effects of marsh vegetation on the accuracy of the LiDAR datasets, finalize a SWMM model for the Drakes Island basin, and develop a beta version of the SMART tool extension.

Tasks to Meet Objectives

• Develop method for accurately extracting vegetation species from Hyperspectral Datasets
• Continue to analyze the errors caused by salt marsh vegetation on the LiDAR
• Remove pressure transducers from the field before ice sets in for the winter.
• Finalize the SWMM model
• Develop a Beta version of the SMART extension for testing

Work Plan for Next Reporting Period
Brown University researchers will use vegetation maps made this summer and last fall and spectral calibration data taken on the day of the over flight to develop a species extraction method from the Hyperspectral imagery. They will investigate the positive or negative differences between using data from the early season flight or the late season flight. Mr. Rogers will be working on the LiDAR datasets to assess the effects of the vegetation on “bare earth” as determined by the sensor. This involves a statistical analysis of ground truth data and vegetation type. Crossover analysis from Hyperspectral data will be very useful if a correction algorithm is developed. Water level monitoring at the Drakes Island culvert will cease sometime in late November/early December when conditions become icy. Mr. Quigley will finalize of the SWMM model for Drakes Island Marsh. The model will be broken into smaller junctions for the purpose of providing better input to the SMART extension. It will also try and incorporate some ground water inputs based on monitoring data collected by US Fish and Wildlife. Upon completion and calibration SWMM will be evaluated for any potential changes to assist in future and more accurate modeling of salt marsh environments. Software Development will continue into the next reporting period culminating with a beta version of the software. Dr. Konisky will alter the software to use model inputs in ARC GIS shapefile format such as water level, elevations, and plant cover from the study site. Input data will be accessed by the SMART model extension and rendered in model-readable formats to analyze baseline conditions at the study site. In addition model processing, logic will be translated from the Visual FoxPro to the Visual Basic language and compiled as a processing module of the SMART extension. A time-series model based on baseline conditions will be generated by the SMART extension to show expected habitat conditions at the site under current hydrologic conditions. The time-series spatial results will be tested to validate model operations under baseline conditions. Support for user-defined ‘what-if’ modeling exercises for Scenario modeling will be added by accepting input requirements.

Anticipated Success in Meeting Project Objectives
The project team has made significant progress in achieving the desired results within the proposed budget and scheduled timeframe. It is anticipated that all objectives will be met for the next progress period. It is too early to tell if Hyperspectral will be successful as anticipated since imagery has just been collected and processed to a point where spectral analysis can begin. We have no reason to believe that the objective can not be met or that the problems with the data collection will hamper the task success.

Overall Project Timeline Update
We are on target with the project objectives and timeline. We do not expect any delays in the proposed timeline.

Preliminary Data
Processing of LiDAR datasets has revealed that marsh vegetation may appear to have an effect on bare earth calculations. This is in support of our hypothesis and seems to be dependant on vegetation type. Vertical profiles through various types of vegetation were taken to help determine light penetration and vegetation density (Figure 3 and Figure 4). SWMM models run to date suggest that the new culvert (Figure 5) at Drakes Island continues to be a undersized restriction. Significant hydraulic losses were observed on both the downstream and upstream culvert orifices. These observations have opened a discussion of how flow efficiency can be improved. Also, as observed in other locations, the SRT is under performing, reversing the spring/neap high tides and allowing higher than permitted water levels to occur on the interior. These events are observed in the monitoring data but are difficult to model due to their relationship with the hydrodynamic force on the gate and wind conditions. A preliminary software tool has been developed as an insert to the SMART extension. This tool called the inundation viewer uses water level data from pressure transducers, NOAA predicted water levels or SWMM export format to visually display water levels in the marsh. We have felt for years that this would be a useful tool for displaying at public stakeholder meetings the effects of current and proposed conditions in and around the marsh proposed for restoration to help achieve project support.

Dissemination
Contact with End Users:
Several conversations that updated the End User as to project status.

Expenditures
Expenditures are within the anticipated range.

End User Advisor Feedback
Name: Tim Smith
Organization: Massachusetts Coastal Zone Management Wetland Restoration Program
Location: 251 Causeway St., Suite 800, Boston, MA 02114
Phone number: 617-626-1247
E-mail: tim.smith@state.ma.us

1) At this stage, what are the potential applications for this research? Please discuss how you and others could potentially use the technology.
The research will result in an invaluable tool for the feasibility assessment, design, and evaluation of salt marsh restoration projects. The tool will aid in preparing grant and permit applications and educating the general public about the effects and benefits of proposed restoration actions.

2) What, if anything, has changed about this project's potential applicability since the last reporting period (not applicable to the first Progress Report)?
Nothing has changed since the last reporting period. MA CZM still feels that this research has significant potential as planning tools for salt marsh restoration tools.

3) Do you see any key challenges that the researchers may want to address or keep in mind?
MA CZM is particularly interested in developing useful and efficient techniques for using LIDAR data for salt marsh restoration design and monitoring and other coastal management applications. Special attention should be paid to ensuring the quality and suitability of LIDAR data.

4) Does this report offer you enough information to adequately address the above questions?
Yes