Kernel South America is Live Learn More

August 05, 2016

In our last post,Desert jewels, we covered some of the basics of using spaceborne observations to monitor plant health and maturity.  Images from 700 km above the Earth, while often beautiful, are a little hard to connect with for those of us who have a more ground-bound profile.  So we started asking ourselves whether we could do something here on the surface that would help us build some intuition for multispectral imagery.

And so TellusCam was born.  The image above is from an inaugural scouting mission through the neighborhood of East Arlington here in Boston.  It turns out that all of our digital cameras have a lot of sensitivity to the same sort of near infrared light (NIR) that made the Landsat image ofSaudi Arabian agriculture possible in our last post.  Camera manufacturers filter that NIR light out as noise, but there arenumerous outfits andDIY manuals who will happily help you strip those filters away.  We took one up on the offering, and now the team is exploring the beauty of false color infrared imagery at human scales.

In the scene above, healthy grass and leaves are reflecting strongly in NIR, so they appear glowingly bright; while the concrete and asphalt absorbs that NIR light, making for a dark backdrop.  This image was shot just after sunset in very low light conditions; essentially the illumination source was the NIR coming from a hazy, overcast evening.  TellusCam also revealed some cool high altitude cloud structures that were completely washed out using to the unaided human eye.  The truth is, there is a lot ofvariation in the way we perceive color and light… among animals, and even among humans.  For a lot of us, the beauty of an orbital perspective is what drew us into the field of remote sensing in the first place; TellusCam is a reminder that science imagery need not be 700 km removed to trigger surprise and wonder.

Also in News

Space….the final frontier! (for crop maps)
Space….the final frontier! (for crop maps)

February 16, 2018

In order to make our pioneering forecasts, we need to know where crops are, which is why we’re pioneering new techniques for crop mapping.

Backtesting To The Future
Backtesting To The Future

February 02, 2018

If you built a machine that predicted the future, how would you test whether it worked? Ideally, you’d be able to get in your time machine and see whether your predictions were right. Sadly, our team includes no MartyMcFlys or Doc Browns... so we use backtesting!
TellusLabs releases three new yield models for South America
TellusLabs releases three new yield models for South America

January 25, 2018

TellusLabs launches three models for South America: Argentina Soy, Argentina Corn, and Brazil Soy.