Using an NDVI Sensor to Estimate Cover Crop Nitrogen Content

Cover crops are an important tool for managing nitrogen (N) in agricultural systems.  New decision support tools are being developed to credit the N supplied or tied up by cover crops and an important input to these tools is the N contained in the aboveground cover crop biomass.  Quantifying cover crop N content has traditionally been done by cutting a biomass sample from a known area, drying it, weighing it, and submitting the dried tissue to a lab to measure the N concentration.  This method, while highly accurate, is tedious, time consuming, and expensive.  To facilitate an easier measurement of cover crop biomass N content, Penn State Extension’s Soil Fertility and Nutrient Management program has developed a method to use an NDVI sensor to scan the cover crop canopy and convert the NDVI reading into an estimate of the biomass N content.

NDVI (normalized difference vegetation index) sensors measure a ratio of the red and near infrared light reflected by a crop canopy.  The NDVI scale runs from 0 to 1, with values closer to 1 indicating greater absorption of red light by the chlorophyll in plant leaves.  These sensors were originally designed to measure the N status of cereal crops to more precisely determine the N fertilizer requirement. Two types of NDVI sensors are commercially available with intended uses in agriculture: i) a handheld sensor that can be walked across a field and ii) a sensor that can be mounted to a vehicle or spray boom and driven across a field.  Both of these sensor types are distributed by Trimble under the Greenseeker tradename.  The handheld type can be used to obtain an NDVI reading of a single point, or the average NDVI value of a transect.  The vehicle mounted type is designed for data logging with a GPS in order to make a map of NDVI over an area or to vary the rate of fertilizer applied on-the-go based on canopy NDVI readings. NDVI can also be measured by a multi-spectral drone camera or satellite imagery.  

This article will focus on how to use the handheld Greenseeker sensor to obtain an NDVI measurement of the cover crop canopy at a point or along a transect and use it to predict cover crop N content.

1. Hold the Greenseeker sensor to the side of the body at arm’s length, approximately three feet above the top of the cover crop canopy.  Orient the sensor so the nose is pointed in the direction of walking and the LED lights and reflectance sensor are pointed straight down at the cover crop canopy.
2. Pull the trigger on the handle and the sensor will start flashing red and infrared LED lights onto the cover crop canopy.  The digital display on the top of the sensor will read the NDVI measurement several times a second.  Once the trigger is released, the average NDVI value measured while the trigger was pulled will be displayed for a few seconds.  Remember this number or write it down, as there is no way to retrieve it once the display goes blank.
3. If you want to take a point measurement, depress the trigger for several seconds while holding the sensor over the point of interest.  Release the trigger to obtain the average NDVI value.  If you want to obtain a transect measurement, hold the trigger down while walking the transect.  After the trigger is released, the screen will display the average NDVI value measured over the transect.  The sensor will only average measurements taken in the first 60 seconds of the trigger pull.  While you are walking, be sure the sensor is held as far to the side of your body as possible to prevent the sensor from reading clothing or body parts.
4. Convert the NDVI reading to lbs N per acre using Table 1.  In the table, NDVI values measured by the sensor are listed in the rows and there are 5 columns of cover crop N content predictions based on different cover crop type by season categories.

Additional Notes on Using the Handheld Greenseeker

• The Greenseeker sensor is an “active sensor” meaning it supplies its own source of red and near infrared light.  Therefore, the sensor can be used in any light conditions any time of day, including full sun, full shade, cloudy, and partly cloudy conditions.  Furthermore, dew on the plant foliage does not appear to interfere with sensor readings in our experience.
• The NDVI relationship with cover crop N content depends on the sensor being able to see the green leaves of the canopy without obstructions above the canopy.  Anything obscuring the green leaves or presenting a different color above the canopy, including flowers of the cover crop, grain heads, and winterkilled residues, will interfere with the sensor reading and generate an incorrect prediction of cover crop N content.
• The error of the sensor in predicting cover crop N content increases with the NDVI reading.  At the highest NDVI readings, the actual cover crop N content will be within 20 lbs N/ac of the predicted value approximately 75% of the time.  Furthermore, the sensor calibration is more likely to underpredict N content than overpredict it.
• Occasionally the sensor display will show an error code rather than the NDVI.  Most often this is because the sensor is either too close, or too far away, from the canopy.  If an error code occurs, adjust the sensor distance from the canopy to between two and four feet.  If this does not correct the problem, consult the user manual to determine if the error code is due to other issues.

In the table below, find the row that corresponds to the NDVI measurement from the sensor.  Then, to determine the predicted cover crop biomass N content, move across to the column that corresponds to the appropriate season and cover crop type for your measurement.  Brassica cover crops include forage radish and canola in fall and canola in spring.  Vining legumes include hairy vetch and peas.  Mixtures can be composed of any combination of brassicas, grasses, or legume cover crop types.