Case Study

Milgadara is a thriving mixed-farming enterprise on the outskirts of Young, NSW. It is owned and run by Rhona and Bill Daley as a multi generation family farm. They began a journey into what they call biological agriculture over 20 years ago, when they experimented with composting and mineral blend soil amendments. They prepare and distribute soil amendments through YLAD organics, and host annual field days to demonstrate best practices to growers in the area. Below shows details of their mixed (cover)cropping and grazing paddocks and how the management has changed with long term crop and pasture rotations over the observational years (2017-2025).

Read more the Milgadara Property as a foundational Soil For Life Case Study Farm here.

Demonstrating PaddockTS:

We ran PaddockTS using mode1 and mode2 (see About) for a 6x6km region around the Milgadara property.

Paddock annotation data was available specifying annual crop type, records of the application of biological/mineral supplements, as well as paddock-level yields.

Below, we demonstrate how PaddockTS results deliver unique insights into how climate and management influence agro-ecosystem outcomes.

Automatically delineate paddock boundaries

Auto-generated paddock boundaries using PaddockTS

User-provided paddock boundaries from Agriwebb (beta farm)

Visualise dynamic landscape history

MILGADARA_b03_2018-2024_manpad_RGB.mp4

Time lapse videos made using Sentinel-2 satellite data give a window into how a landscape has transformed. On the left is a "visual" RGB representation of the reflectance data.

These videos give users a broad lay of the land that can complement extensive on-ground management, yield and phenology data. In addition, users can compare this top down historical record of landscape change with the preceding weather data, like that shown below; and also gauge differences across a broad landscape.

MILGADARA_b03_2018-2024_manpad_vegfrac.mp4

Above video displays a "false-colour" representation of landscape cover change, based on a spectral model that predicts the fractional area in each 10m pixel of photosyntethic vegetation (green), non-photosynthetic vegetation (blue) and bare ground (red). So, bright green represents a high fraction of photosynthesising plants, while, blue shows brown or dry biomass, and red is the signal of bare ground stage after tilling, heavy grazing or prolonged drought.

Summarise weather data

PaddockTS generates a summary of historical environmental variation at the site including rainfall, soil moisture, temperature and evapotranspiration (ET). Grey arrows in the top panel show the dates when Sentinel-2 satellites passed over the site and collected light reflectance data.

Paddock "calendar view" lays out satellite imagery of a selected paddock in a grid of weeks (columns) by years (rows).

An intuitive snapshot of how specific paddocks changed within and between years, with different "green-up" and "brown-off" times for the main winter growing season, and occasional summer growth.

In this paddock:

Effects of extreme drought building through the summer of 2018-early 2020 are evident
After the 'tinderbox' drought broke, longer growing seasons followed from generally higher-rainfall years.
Crop greenness and potential growth can be traced from green up after sowing through the winter growing season and finally to harvest in early Summer.
A 2021 canola crop can be seen flowering in Spring. Confirmed by the management label for this paddock provided through Agriwebb (public beta farm)
Paddock "No. 4" management was split in later years which aligns PaddockTS segmention into 2 paddocks #27 and #32 above.

Compare seasonal dynamics of greenness and yellowness for different years and crops across paddocks

These plots trace two useful vegetation indices for a selection of Milgadara paddocks in a dry year (2019) and a wet year (2021).

The normalised difference vegetaiton index (NDVI) is a widely used greenness metric reflecting chlorophyll levels and leaf area in plant tissue. Green leaf area is necessary for photosynthesis under sufficient light and water conditions, leading to biomass growth and seed yield.

Here we see:

greater overall growth in a year with higher rainfall and soil moisture.
Summer cover crops for certain paddock-years that are thought to have a beneficial effect on productivity later in the season.
paddocks used for intermittent grazing (Ranch and No 1), follow quite different dynamics to cropping paddocks, likely modulated in part by the intensity of grazing.
tree rows exhibit distinctly muted variation in photosynthesis within and between years

The canola flowering index (CFI) is a metric designed to detect higher content of pigmented tissue in canola flowers, and is a clear indicator of crop types. Flowering time is a key developmental stage, under genetic and environmental control.

Here we see:

Phoenix canola variety flowered later and more intensely than the TT Y44 variety, with possible impacts on yield.
A 2019 canola crop (Rocky East paddock) that was grazed during the winter did not show strong flowering which decreased seed yield.

Overall, these this demonstrate how paddock time series (PaddockTS) contain the signatures of the biological and managed processes that drive agroecosystems.

These croptype and growing season observations for each paddock year are the necessary input to run biomass, development and yield growth models, like our Dynamic Agro-Ecosystem Simulator. DAESim uses the daily weather and soil moisture, crop type, planting and harvest date as inputs, to simulate photosynthesis, transpiration, carbon allocation, and leaf, root, stem and grain biomass growth as outputs.

Paddock TS with DAESim enable accurate estimation of the agricultural carbon cycle.

Page updated

Google Sites

Report abuse