Estimating Productivity on the UBC Farm with Remote Sensing

This research was based in the traditional, unceded, and stolen land of the Coast Salish peoples, including the territories of the xʷməθkwəy̓əm (Musqueam), Skwxwú7mesh (Squamish), Stó:lō and Səl̓ílwətaʔ/Selilwitulh (Tsleil-Waututh) Nations.

This research was conducted on the ancestral lands of the Attawandaron/Chonnonton, the Anishinaabe, and Haudenosaunee peoples and the treaty lands and territory of the Mississaugas of the Credit. Guelph is in the "Dish With One Spoon Territory,' signifying a treaty between the Anishinaabe, Mississaugas and Haudenosaunee that bound them to share the territory and protect the land.

The indigenous peoples of these lands were, and continue to be, the original stewards of the soil and earth on which this research is based.

We are all familiar with the idiom that the grass is greener on the other side of the fence. At the UBC Farm, we wanted to know if the grass (or crops, in this instance) really were greener in some areas of the farm, and if so, by how much.

Ecological agriculture promotes soil and plant biodiversity by planting a wide range of plants together (what's known as polyculture) as well as over time (what's known as crop rotation and cover crops). These agricultural practices capitalize on nature's inherent strengths, and reject chemical forms of soil amendments. Resilient and productive soil is a farmer's greatest resource, and farmers often 'feed the soil' to promote health and productivity in their crops. In ecological agriculture, how much a plant grows is a strong indicator of a large assortment of metrics together known as 'soil health.'

Finding efficient and effective ways to monitor soil health is therefore incredibly important to farmers. Recent advances in satellite imagery technology have allowed for monitoring large farms from space for areas that are experiencing nutrient deficiency, water stress, or pests. This has contributed to greater yields and more efficient use of resources, since reactions to any problems can be very accurately (and quickly!) targeted.

However, not all farms have benefited equally from this technology. Small-scale farms under diversified production make up a significant proportion of global agricultural lands. Because of their small size, and crop diversity, there are several obstacles to overcome for satellite imagery to be a useful tool for these farms. This research, in partnership with the UBC Farm, was designed to explore the potential of studying soil health on the UBC Farm from space.

The UBC Farm is a 24-hectare integrated farm, forest, and community learning space located at the University of British Columbia (UBC), in Vancouver, British Columbia.

The farm is incredibly diverse in both its composition and its functionality: half of the land is currently under native and regenerated forest, and approximately a third is in active food production, dispersed over 27 crop cultivation plots. These plots are a part of either an eight or four year rotation system, in which the same family of plants is not grown in the same area for that many years. This extensive effort is maintained as a means to improve soil fertility, plant productivity, and soil biodiversity.

The remaining land on the farm is divided into perennial hedgerows and orchards, pasture for laying hens, Indigenous community space, bee hives, and teaching gardens.

For this study, researchers collected 12 monthly satellite images from 2019 (with the exception of one image from January 1, 2020 because a suitable image for December 2019 was unavailable). These images were used to estimate greenness and vegetative health on the Farm.

Vegetative health and greenness were then used as a proxy for productivity, and in turn, soil health.

The satellite imagery collected captured significant differences in greenness throughout the year. October was found to be the most productive, and June was found to be the least. The differences can be seen in the two images below.

There were only minimal variations in greenness observed across the different farm plots over the whole year, meaning that productivity and greenness were distributed relatively evenly across the fields.

The results demonstrated a clear seasonal trend in greenness across the farm, but they likely represent a different season than we were expecting to see.

There are three main aspects to consider to understand these surprising results. Firstly, growth on the farm in the spring and summer is dynamic with plots continually being replanted and harvested at irregular, but frequent, rates. Secondly, the farm plants crops known as 'cover crops' over the winter months. The cover crops are planted once, and then not harvested until shortly before the farm is preparing the land for spring planting. This means that over the winter months much of the farm is consistently covered in vegetation. Lastly, the images captured represent a brief snapshot of a busy farm in any given month - they do not capture all of the growth from that month.

With these three considerations in mind, the low greenness values in the summer and the higher values in the late fall/winter most likely speak to level of activity on the farm. When the farm is more active in the summer months, there is a greater chance that the snapshot would capture a day where much of the farm had recently been harvested or replanted. When the farm is less active in the late fall and winter months, on the other hand, the majority of what is planted is static cover crops and any single snapshot will appear much more green.

This study was helpful in illustrating what adaptations to how we used the technology might provide more useful results for small-scale farms like the UBC Farm to integrate into their farm management plans. Increasing the number of images taken each month as well as the detail captured in each would improve the depth of analysis we would be able to do, and with it the results we could provide to the farm. New advancements in collecting 3D data of the farm would also allow us to measure indicators of soil health other than greenness.

This study fits into an exciting and novel field of research looking to equalize the playing fields between large and small farms in how they can make use of new and rapidly innovating technology. We hope this research accelerates growth both in future research and in the field.