3

I have a study area in N. Florida loblolly pine plantation with 3 treatments (drought, fertilization, drought+fertilization, control) and 4 replicate plots. All plots will be in the same scene. These treatments are expected to alter the growth and response of pine trees.

The only available existing imagery at the initial time of treatment installation is from quickbird (one in March and one in December) - this will be the baseline image. Ideally, it would be nice to acquire imagery every year and season, it's not feasible due to budget constraints. So, I will probably purchase one, maybe two more images over the next 1-2 years. They should be the same season as the "baseline" scene.

My Question: Which season, if any, is best to discriminate spectral signatures between treatments/plots to really tease out differences in growth due to treatments and estimate NDVI, LAI, and other vegetation indices? I would suspect winter, when everything is not as lush and green and somewhat stressed (relative to summer). What is your expert opinion? Thanks in advance.

Improve this question
4
  • 2
    If you are interested in measuring vegetation vigor, consider looking into Worldview-2 imagery from Digital Globe digitalglobe.com/…. With these data you will be able to utilize the red edge and yellow bands for veg analysis. Have you looked into free 4-band 1m NAIP imagery from the NRCS Geospatial Data Gateway? Seeing that you are in Florida, the season is not as important as the spectral bands.
    – Aaron ♦
    Commented May 13, 2013 at 14:38
  • 1
    I have not looked into WV 2 or NAIP imagery, but will give it a go. I agree, spectral bands will be most important, but season will surely have an effect as H20 content in leaves and plant vigor can have a big effect on reflectance. The reason for going with quickbird is it is the only imagery available at the time of treatment installation. I suppose baseline image is not as important as what the plot will look like in two or three years, but it would be nice to have a baseline image to make comparisons (before and after) - with the same imagery in the same season. Thanks for the suggestions!
    – derelict
    Commented May 13, 2013 at 14:57
  • Given the budget allows it, I second the advice to look at WorldView-2 imagery. The 8 multi-spectral bands can be a great deal in multi-spectral analysis.
    – Nikos Alexandris
    Commented Oct 12, 2013 at 19:18
  • Thanks for the comments, but they don't address the question. I need to establish a baseline - imagery immediately before the the treatments were applied. This limits me to Quickbird because it's the only available imagery in that area for that time.
    – derelict
    Commented Oct 18, 2013 at 17:01

2 Answers 2

Reset to default
1

Sometimes the season you choose for your imagery analysis is limited by the availability of data due to cloud cover. This is especially true at lower latitudes with a rainy season and dry season. I would suggest starting your search by seeing what cloud free imagery is actually available as that may be the deciding factor of what season you use.

Improve this answer
1

It really depends on the location of your study region and your own research goals. I have experience working with aerial photography collections and typically the ideal times to collect aerial data for vegetation analysis are October/November and February/March. Images taken in October before the leaves have started to change are ideal for measuring LAI, quantifying tree crown size, and identifying tree species. Images taken in February/March are better at evaluating overall ecosystem structure, tree dispersion patterns etc. because the canopy is less dense.

I could give a better recommendation if I had more information about your plots (sizes, number of trees etc.) and your specific research goals.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.