Remote Sensing Q & A
What can we see with satellite imagery?
The satellite
imagery we offer is composites of near-infrared (NIR), red, and green
light. NIR, in particular, is quite valuable to agriculture, as it is
reflected more by the cell walls in vegetation than by soil or water:
the denser the canopy of vegetation, the more NIR light is reflected
back to the satellite. This means that an NIR image of your field
shows the areas of high and low vegetation density with ease, allowing
you to determine the productivity of those areas.
How accurate is the satellite imagery? How detailed?
Our satellite imagery is all radiometrically and geometrically
corrected, so that the georeferencing is quite accurate.
As
far as detail is concerned, we offer various kinds of satellite
imagery, but our most plentiful supply of images is from NASA's
Landsat 5, Landsat 7, and Terra(ASTER) satellites. When processed,
this imagery has a resolution of 5 meters, or around 16 and a half
feet. This means that you can differentiate between varying levels of
vegetation down to 1/160th of an acre, which gives you a highly
accurate picture of your field's variations.
Can I use satellite imagery to determine the cause of the
variation?
No, not by itself. You will still need to "ground
truth" the areas of high or low density; remote sensing is not a
substitute for agronomic knowledge. Where the imagery is helpful is
in showing you where you need to investigate. In other words, it's an
additional source of data, not a replacement for other information.
Okay, so now that I know how the density of vegetation varies in
my field, what can I do with that information?
The satellite
imagery, when analyzed, gives you georeferenced management zones
within the field. This can be helpful for crop scouting, by showing
you the areas you need to inspect, as well as setting yield goals,
determining trends over several years, and not least of all, targeting
the use of seed, water, fertilizer, pest control products, and other
resources to the areas where they would do most good.
Can I use image analysis to make yield predictions?
It
can be helpful as a source of data for such predictions, but you still
need to know your fields. High-density vegetation is often healthier
vegetation, but it may not be: if you plant your crops too close
together, the satellite will show high vegetation density, but the
yield will be poor compared to crops that are appropriately
spaced.
Also, remember that a satellite image is just a
snapshot of the field at the time the image is taken: subsequent
changes, such as weather or disease, can have an impact that the image
won't show (although a later image may well show the results).
What about precision agriculture? Can satellite imagery be
useful for variable-rate farming?
Of course. In fact, that is
one of the primary uses for satellite image analysis: creating
management zones with which to vary the application rate of various
products, such as seed, water, fertilizer, pesticide, and so on.
Exactly how you do this will depend upon how the zones correlate to
the characteristics of the field, such as soil type, moisture,
topography, and the like, as well as to your yield goals.
You
may, for instance, decide to concentrate your fertilizer application
in the high-yielding areas of the field, while leaving the lower-yield
areas with less nutrients, if those areas have relatively poor soil.
Or, in another case, the low-yield areas may well have good soil, but
insufficient water; more irrigation would improve the situation. Or
you may discover that poorly-performing zones are infested with a
disease, and use this information to help stop it from spreading.
How often should I analyze a satellite image?
This
depends on a number of things: what crop you have, what stage of
growth it's in, what you are looking for, and how often you expect
changes to occur. Sensitive crops, or ones of high value, such as
cotton, potatoes, or vegetables, may need to be checked biweekly, or
even weekly. Others may only need to be checked once a month.
Of course, if that month saw a large hailstorm blow through and
destroy large areas of your crop, it may well be useful to check
another image much sooner, in order to assess the damage.
What kinds of analysis can I do on your sites?
We currently offer three types of analysis:
- Near Infrared: A straight measurement of the NIR reflection from the image. This gives you the variations in vegetation, soil type, and other influences. Good for showing variations during peak growth periods.
- Normalized Difference Vegetation Index (NDVI) - Red Band: By calculating the contrast between reflected NIR light, and red light, this analysis can compensate for atmospheric interference in the image. This provides a more averaged view of vegetation density patterns throughout the growing season.
- Normalized Difference Vegetation index (NDVI) - Green Band: A variation on standard NDVI, this analysis contrasts the NIR reflection with green light, which can be helpful for measuring nitrogen content, and thus the health of the vegetation.