Remote Sensing of Large Watershed Change for ecosystem management

Year 2009
Proponent G.A Borstad Associates Ltd. (08 LR 86)
Project type Habitat
FSWP funding source Living Rivers
Grant amount $46,000
Total project value $59,145
ID number 09-LR96-H

Location: Basin wide

Project Summary

Overview

NDVI is an established indicator of vegetation presence, amount and condition frequently used in forestry, land use, agriculture, and studies of climate change1.  In the first year of this project, we obtained the Canada Centre for Remote Sensing AVHRR NDVI dataset for British Columbia for the period from 1985 to 2006, at 10-day and 1-km resolution.  We partially corrected data calibration, filtered for cloud and snow, and extracted spatially averaged time series for selected Fraser watersheds.  We then developed a set of experimental, derived secondary indicators, including seasonal and annual anomalies from long-term NDVI averages, annual amplitude, and timing of spring Ōgreen-upÕ.  In 2009, we will compare these watershed indicators with DFO salmon survival data assembled this year, and work with John Reynolds group to begin investigation of linkages to their more local indicators.  We will also generate image versions of selected indicators for use by other groups (e.g. NCC, ESSA Technologies, Simon Fraser, Upper Fraser Conservation Alliance).  We will participate in the Annual FSWP Meetings to broadcast our results and seek further linkages.


1Myneni et al, 1997.  Increased plant growth in the northern high latitudes from 1981 to 1991.  Nature 386: 698-702

Objectives
1.Data checks: Complete data quality checks on the NDVI dataset, and update the time series to 2008 if data is available.
2. Compare indicators: Relate NDVI secondary indicators to salmon survival in watersheds of interest to DFO, and begin investigation of linkages to local ecosystem indicators generated by John ReynoldÕs group at SFU.
3. Generate products: Generate images of secondary indicators for wider use by other projects (eg. Upper Fraser Conservation Alliance), and for use in outreach.
4. Outreach: Participate in meetings and public fora in order to demonstrate our products and make further linkages with other projects.

Methods

Objective #1:  We will obtain the most recent imagery available from CCRS and append it to the existing time series, making calibration adjustments as required.  Our NDVI time series will be compared to other vegetation indices prepared independently in order to check the calibration.

Objective #2:  Fish survival statistics obtained from DFO and others will be compared with NDVI-based indicators on a year-by-year basis using a correlational approach at the watershed level.  Where appropriate, we will also perform pixel-by-pixel, image level correlations in order to generate maps of statistical probability that can be used to locate geographic areas most highly correlated with the survival of individual fish populations.  Our experience in other projects indicates that these correlation maps are powerful tools to generate new questions and insights.

Objective #3:  The secondary indicators developed in the first year of the project at the watershed level will be applied to the NDVI time series at the pixel level to produce images of secondary indicators.  This will involve the development of custom software.  We will interact with other groups such as the NCC, the Reynolds group and the Upper Fraser Conservation Alliance to customize our products for their use.

Objective #4:  We will attempt to produce intuitive, visually appealing outreach products aimed at raising public awareness of watershed health.  We will publicize the results of our work at the Annual Fraser Salmon Watersheds Meeting and/or other appropriate fora, with oral and poster presentations and attempt to forge further links with other salmon projects. 


Final Results

The final report for the study is posted in the orange resource box to the upper right.


Name:

Email:

Location:

URL:

Comment:

Remember my personal information

Notify me of follow-up comments?

Comments

Be the first to leave a comment on this page!