Pretty picture or valuable management tool? Validation is key!


by James Taylor You hear a lot about canopy sensing nowadays and this magical term NDVI – Normalized Difference Vegetation Index – which is just a standard ratio between how much of the red and the near infrared light in sunlight that leaves absorb (or reflect). In general terms a happy healthy leaf absorbs lots of red light for photosynthesis and reflects lots of near infrared light. A sickly leaf is not photosynthesizing so it reflects more red light and the poor leaf condition also traps more near infrared light, so sick leaves have lower near infrared reflectance. These differences allow canopy sensors to be used to map whether an area of vineyard has happy, healthy vines or not so happy vines. However, the difference in the canopy response (e.g. NDVI) is a relative value. So it tells me that this area has a higher value than that area, but it can’t directly tell me if the level of ‘happiness’ of the vines is actually limiting production.

So after scanning with a canopy sensor and making a map (see the recent blog on this), what could a grower do with this information? Well if you do nothing, all you have is a pretty picture. What you really want to be able to do is convert the colors into a value that means something for management, such as pruning weight or vine size.

Unfortunately there is no magic wand or silver bullet to do this. Every vineyard is different and every canopy sensing survey is different, so for each survey all those NDVI values are affected by many different things, e.g. the method of sensor mounting, the time of survey, pruning practices etc… but thankfully for us, most of these things are usually standard in a vineyard block, so higher NDVI values should always indicate happier vines and lower NDVI values less happy vines. The bad news is that if you want to convert those pretty colors into pruning weight values you will have to take some pruning weight measurements to calibrate the sensor data. There is no way around this at the moment. The good news is that with the canopy maps, the LERGP team has a simple protocol to help growers achieve this. The protocol is specifically designed to make use of the NDVI maps to minimize the effort required and maximize the value of the information collected.

The protocol is being published in the new Catalyst Industry Journal (  which, for those of you who haven’t heard about it, is a new journal aimed at growers and viticulturists to clearly explain advances in viticulture science. It is written with end-users, not scientists, in mind. The protocol is also available from LERGP if you can’t wait (or don’t have a subscription).

Anyway, the protocol has a few simple steps to follow to get good pruning weight maps from canopy sensing and while I can’t explain it in detail here, it essentially

  1. Provides instructions for setting up sensors properly in vineyards

  2. Uses the canopy vigor maps to be smart about where we sample along a few dedicated rows.

  3. Uses panel length weights not individual vines to minimize the time needed to sample in the field.

  4. Provides a method for a grower or service/extension provider to link the canopy sensor map with the pruning weight samples to get a pruning weight map of the entire vineyard.

The process is summarized in the figure below. Ideally, as we move forward we would like to make this a more automated process to make it even easier for growers to adopt. NDVI or canopy vigor maps are only a stepping stone and until we take further steps we won’t unlock the value in this information.