arrow right cross menu search arrow-left arrow-right menu open menu close plus minus

Recorded Webinar: How Remote Sensing Powers Agriculture

Remote sensing measures crop growth throughout the year so that the growers can analyze conditions based on the data and take action that will have a positive influence on the harvest outcome. But what exactly is remote sensing?

Watch our recorded webinar to hear from members of CIBO’s science team including Ernesto Brau, Lead Computer Vision Scientist, Ph.D., Shane Bussmann, Senior Data Scientist, Ph.D., and Pankaj Bhambhani, Data Scientist. The CIBO team will share an overview of remote sensing, ways to use the data and how remote sensing data can be used for decision-making among the agriculture and food security communities.

Here are some of the questions this webinar answers 

Why is remote sensing important to agriculture?

For agriculture, it’s so useful and valuable because we’re able to observe large pieces of farmland. The satellites are over the earth and give us a picture of what is happening across the US. We’re able to observe all the fields in the US over time to see what’s changing and what isn’t changing. No other tool lets us monitor or scale to such a vast amount of area.

What’s the difference between a drone and a satellite? 

One of the main differences is scale. A drone has a limited spatial scale. The other difference is that satellites have much higher quality imagery. The spectral characteristics could also be quite different.

With a satellite that we get is consistency in the measurement and calibration. We get the whole US at the same calibration. With drones, you might have two different properties depending on the drone. Even one drone can have different qualities. For example, how bright a crop looks, how green vs brown. At CIBO, we measure the number there but any difference will impact the result from the algorithm.

There is whole field in computer vision dedicated just to color consistency. For example, white can look vastly different in different conditions.

You also need to take into account that the light from the sun will be changing over time. Atmospheric correction cannot be applied to drones. Satellites do the correction on their images. For us, using drone data is difficult because we don’t know if we can trust the data because it may not have been cleaned correctly.

What data sources are used for remote sensing? 

There are many sources of data that are easily available to analyze. At CIBO, we mainly use images from passive remote sensing. We use satellites that reflect light from the sun and it records the light. They are global observations about land.

There are two main sources – Landsat and Sentinel. They are easily available and high quality.

We don’t own satellites. We use satellites that are owned and operated by the government. The Landsat satellite was launched in 1972 so we have almost 50 years of data. There have been nine Landsat missions which we pull our data from.

The Sentinel satellites are run by Europe. They have been live a shorter time but they have slightly better quality data. The main advantage of using these satellites is that they are free and provide global coverage. Private satellites can be very expensive and more limited in scope.

A proper calibration of the data must be done. The raw remote sensing data cannot be used. You need standardized pipelines in place to clean the data and make it usable. Finally, they also provide good quality supplementary products. They have flag where there may be clouds and help flag areas that may be covered by a cloud and unusable.

When we think about what we like about the government data, the pricing is a key bonus. It makes it a lot more feasible to use and it doesn’t break the bank. The other component is that we are able to trust the data. An image on April 15 is just as trusted as May 15. You see corn starting to pop out of the ground and vegetation appear. You can trust that crops are really green and you don’t need to worry about the color calibration.

What are the limitations of remote sensing? 

There are limitations. With the Landsat and Sentinel satellites, they are orbiting the earth. In this case, we get images 1-4 times a week. However, oftentimes there are clouds. You can go a whole month without seeing the land because of clouds covering the area. We work around that by taking a look at what was happening with a neighbors field or similar areas.

If there’s something that only happens in an hour or ½ hour, you might want to see someone tilling their field but you won’t see it because you’re only getting two observations a week.

Watch the Webinar