Yield, NDVI and EM layers show pattern; named points show mechanism. How Australian precision growers keep both honest.
Yield maps, NDVI layers and electromagnetic surveys are seductive. They paint the paddock in colour, smooth out the noise of memory, and let you argue with your neighbour using something other than hand gestures. For Australian growers juggling variable rainfall, striping from old machinery, and soil types that change inside a single pass, those layers can feel like finally seeing the truth.
They are not the truth by themselves. They are summaries of what happened — biomass, moisture proxies, conductivity patterns — stretched across space and time. Portable soil screening at named points is a different kind of summary: what the soil is doing right now at places you can find again. The productive conversation happens when you put the layers side by side and resist the urge to merge them into one number that pretends to be precision.
This Field Note is for growers and agronomists who already have yield or imagery in the cab or the office, and who want TerraTactics-style screening records to mean something beside those maps. It is not a prescription for variable-rate everything. It is a way to keep maps honest, zones defensible, and laboratory money pointed at the right questions.
WHAT YIELD AND NDVI ACTUALLY SHOW
A yield map is an accounting document dressed as geography. It records mass flow at harvest — often corrected, sometimes not — and assigns it to coordinates that depend on GPS quality, header width, moisture correction and the mood of the yield monitor on the day. A good yield map reveals stable high and low zones, machinery effects, and seasons when the whole paddock was simply too dry to tell a soil story. A bad yield map is treated like soil fertility when the real driver was frost, waterlogging, rabbits or a one-off spray miss.
NDVI and other vegetation indices compress greenness and biomass into another surface. They respond quickly to nitrogen, moisture, disease, crop stage and soil constraints that limit root exploration. In a wet southern season, NDVI can look magnificent on ground that will collapse under heat in spring. In a northern irrigation block, NDVI can mask salinity at the edge of a pivot because the crop is still green until it is not.
Electromagnetic induction maps and related conductivity layers describe how easily current moves through the profile — a blend of moisture, clay, salt and sometimes compacted structure. They are brilliant for finding where to walk first. They are not a substitute for knowing whether the salt is in the root zone this year, or whether acidity lives ten centimetres below a friendly-looking surface.
Each layer answers a different question. Yield asks what left the paddock. NDVI asks what the canopy looked like when the satellite or sensor passed. EM asks where the profile tends to hold water and salt differently. Screening at a point asks what you would feel and measure if you stood there with a probe at a defined depth on a defined day.
ZONES VERSUS POINTS: WHERE RECORDS BREAK
Precision agriculture loves zones. Management zones aggregate pixels into polygons that get one rate, one variety choice, one narrative. Zones are useful because no farmer can manage every square metre independently. They are dangerous when they are drawn once from one season’s yield and then treated as eternal soil classes.
Points are the opposite discipline. A named point is small on purpose. It is not the paddock average. It is a witness — high on the rise, low in the swale, on the saline shoulder, beside the old fence line where lime went on five years ago. When you screen at points repeatedly, you build a time series that zones alone cannot give you, because zones move when you change crop, tillage or water.
The failure mode is collapsing points into zones too early. Three low-yield years in the eastern third does not prove low pH throughout the third. It proves the eastern third underperformed for reasons that might include soil, might include drainage, might include variety fit. If you lime or gypsum the whole polygon because the map is red, you may spend well on a fraction of the area and miss the one wet corner that was actually the problem.
The healthier pattern is dual bookkeeping. Keep zones for machine files and broad strategy if that is how you farm. Keep named screening points for soil watchfulness. When a zone misbehaves on the yield map, ask which witness points inside that zone saw conductivity or pH move first. When a point misbehaves, ask whether the zone story still holds or whether you are looking at a sub-zone the polygon erased.
PAIRING LAYERS WITH SCREENING WITHOUT FALSE PRECISION
False precision is what happens when a smooth colour ramp implies certainty the underlying data did not earn. Yield at two tonnes per hectare resolution does not mean you know salinity at two tonnes resolution. NDVI at five-metre pixels does not mean pH is uniform inside each pixel. EM at twenty-metre swaths does not mean the same conductivity depth applies under every tree line.
Screening adds precision of a different sort: repeatability at a spot. If Wirraway West 03 reads consistently higher in conductivity than Wirraway West 01 under matched moisture, that is a real difference at those spots. Overlaying that difference on an EM map might confirm clay or salt trend — or might reveal that the EM layer missed a small saline seep because the survey was done in a dry month and the seep only expresses when the watertable rises.
Practical pairing steps:
Walk the map before you walk the paddock. Mark three to six places where layers disagree — high yield but low NDVI late season, high EM but average yield, stable yield but shifting NDVI year to year. Those disagreements are where soil stories hide.
Visit those places as named points, not as “somewhere in the red bit”. Record depth, moisture, crop or pasture, and the seven screening indicators if you use TerraProbe. You are not trying to paint the polygon. You are trying to explain whether soil at witness depth matches what the layer implied.
Hold season and management constant in your head. A yield layer from wheat after wheat tells a different story than wheat after fallow with a legume break. If you compare screening from a legume year to yield from a cereal year without noting the rotation, you will invent soil problems that are really rotation effects.
Use layers to prioritise laboratory sampling, not to replace it. When yield and NDVI both slump in a zone and conductivity screens climb at two witness points, that is a stronger brief for a salinity or sodicity package than either layer alone. When yield is fine but pH screens slide at a rise, the map will not warn you — the points will.
AUSTRALIAN CONTEXT: WHEN MAPS LIE KINDLY OR CRUELLY
In Mediterranean-climate dryland, one heroic season can make poor country look capable on a yield map, and the next drought can make good soil look hopeless. Multi-year yield stacks help, but they still average away the year when deep nitrogen or deep moisture saved a crop on shallow soil. Point screening that survives those seasons shows whether conductivity or pH actually changed, or whether the map only reflected rain.
In irrigation districts, uniformity is the fantasy every map sells. Centre pivots circle over zones that were never level, over sodic tails that only show up when application rates creep, and over bore water that shifts in salinity between summers. Pairing EC screens at fixed points with NDVI inside and outside the wetted circle catches edge effects that yield maps blur because the header still took grain from both.
In pastoral country, yield maps may not exist at all. NDVI or ground cover imagery still partitions the paddock. Screening at watering points, gilgai, and invasion fronts gives a soil record where “production” is wool or cattle weight, not tonnes per hectare. The same zone-versus-point logic applies: a bare patch on imagery is not automatically acidity; walk a named point before you commit lime.
If you advise several farms, say aloud that layers are property-specific. A conductivity colour scale that works on the Liverpool Plains does not transfer to a sand plain near Esperance without re-grounding. Witness points anchored to each property stop you from importing another farm’s zone map as gospel.
TERRATACTICS AS THE GLUE, NOT THE MAP
TerraTactics does not generate yield surfaces or satellite imagery. It holds screening history at named locations, with GPS assistance, moisture and depth notes, and charts that belong to the grower. That is intentional. The value is keeping point records stable while maps rotate through seasons and providers.
Beside a yield map on a laptop, keep the handoff honest: the map suggests where to look; the probe and routine decide what to measure; the Grower portal keeps the series when the map file is archived on a USB stick somewhere in the office. Optional TerraTactics Soil Analysis on Grower plans can summarise screening change — still insight, not a replacement for laboratory chemistry or variable-rate prescriptions.
For farm business records, exporting or simply maintaining a multi-year point history beside map screenshots in your own filing system is enough for many lenders and advisers. The credibility comes from named locations and consistent method, not from claiming the probe mapped the paddock.
COMMON MISTAKES WHEN LAYERS MEET POINTS
Treating one witness point as proof for an entire zone. Points justify investigation and targeted lab samples, not whole-paddock amendments from one reading.
Ignoring moisture when comparing screening to NDVI. A green image after rain can coincide with a conductivity spike that is mostly water, not salt. Note moisture every time.
Redrawing zones every year without updating witness points. If zones move but points stay put, you lose the thread of what changed.
Assuming EM depth matches probe depth. EM integrates through the profile; your standard 100 mm screen is not the same measurement. Use EM to choose where to point the probe, not to infer probe numbers without ground truthing.
Splicing yield from different headers, crops or moisture calibrations into one “master map” and then correlating to pH as if years were interchangeable. Fix the metadata or fix the habit.
Letting software auto-generate management zones without a paddock walk. The walk is where you discover the rock strip, the old burn, the strip of sodicity too narrow for the pixel.
A WORKFLOW FOR A TYPICAL SEASON
Pre-season: open last year’s yield and NDVI if you have them. Mark witness points for zones that underperformed or overperformed unexpectedly. Sample those points under your normal SOP before major inputs.
In-crop: if imagery dips in a zone that screened poorly at sowing, note whether it is uniform or patchy. Patchy often means machinery, pest or local drainage. Uniform often means nutrition or water. Screening mid-season is optional; when you do it, keep depth and moisture honest.
Post-harvest: yield map finalised (with whatever calibration you trust). Compare zone averages to witness point trends. Escalate to laboratory where point trends and maps agree something is wrong, or where they disagree and you need to know which lied.
Between seasons: do not delete points because they were “inconvenient”. Inconvenient points are usually the farm teaching you something.
FOR ADVISERS
When you sit with a grower who owns fancy maps but no point history, start with one paddock and three names. Show how a conductivity screen at a rise explains a chronic low-yield stripe better than another zone redraw. When you sit with a grower who owns probes but no maps, explain that screening without spatial context still beats memory, but layers help prioritise where to spend walking time.
Never sell TerraTactics as a yield-map replacement or a variable-rate engine. Sell it as the record that keeps maps accountable across staff changes and wet years.
BOTTOM LINE
Yield maps, NDVI and EM layers describe outcomes and proxies across space. Named point screening describes soil conditions at places you return to, with method you control. Zones help machines; points help truth. Pair them by letting layers suggest questions and letting repeatable screens answer whether soil at witness depth belongs in the story — then use the laboratory when money and risk demand proper chemistry.
False precision is the enemy. Honest pairing is the habit. Walk the disagreeing pixels, name the spots, screen with discipline, and let the paddock teach you what the colours only hinted at.
They are not the truth by themselves. They are summaries of what happened — biomass, moisture proxies, conductivity patterns — stretched across space and time. Portable soil screening at named points is a different kind of summary: what the soil is doing right now at places you can find again. The productive conversation happens when you put the layers side by side and resist the urge to merge them into one number that pretends to be precision.
This Field Note is for growers and agronomists who already have yield or imagery in the cab or the office, and who want TerraTactics-style screening records to mean something beside those maps. It is not a prescription for variable-rate everything. It is a way to keep maps honest, zones defensible, and laboratory money pointed at the right questions.
WHAT YIELD AND NDVI ACTUALLY SHOW
A yield map is an accounting document dressed as geography. It records mass flow at harvest — often corrected, sometimes not — and assigns it to coordinates that depend on GPS quality, header width, moisture correction and the mood of the yield monitor on the day. A good yield map reveals stable high and low zones, machinery effects, and seasons when the whole paddock was simply too dry to tell a soil story. A bad yield map is treated like soil fertility when the real driver was frost, waterlogging, rabbits or a one-off spray miss.
NDVI and other vegetation indices compress greenness and biomass into another surface. They respond quickly to nitrogen, moisture, disease, crop stage and soil constraints that limit root exploration. In a wet southern season, NDVI can look magnificent on ground that will collapse under heat in spring. In a northern irrigation block, NDVI can mask salinity at the edge of a pivot because the crop is still green until it is not.
Electromagnetic induction maps and related conductivity layers describe how easily current moves through the profile — a blend of moisture, clay, salt and sometimes compacted structure. They are brilliant for finding where to walk first. They are not a substitute for knowing whether the salt is in the root zone this year, or whether acidity lives ten centimetres below a friendly-looking surface.
Each layer answers a different question. Yield asks what left the paddock. NDVI asks what the canopy looked like when the satellite or sensor passed. EM asks where the profile tends to hold water and salt differently. Screening at a point asks what you would feel and measure if you stood there with a probe at a defined depth on a defined day.
ZONES VERSUS POINTS: WHERE RECORDS BREAK
Precision agriculture loves zones. Management zones aggregate pixels into polygons that get one rate, one variety choice, one narrative. Zones are useful because no farmer can manage every square metre independently. They are dangerous when they are drawn once from one season’s yield and then treated as eternal soil classes.
Points are the opposite discipline. A named point is small on purpose. It is not the paddock average. It is a witness — high on the rise, low in the swale, on the saline shoulder, beside the old fence line where lime went on five years ago. When you screen at points repeatedly, you build a time series that zones alone cannot give you, because zones move when you change crop, tillage or water.
The failure mode is collapsing points into zones too early. Three low-yield years in the eastern third does not prove low pH throughout the third. It proves the eastern third underperformed for reasons that might include soil, might include drainage, might include variety fit. If you lime or gypsum the whole polygon because the map is red, you may spend well on a fraction of the area and miss the one wet corner that was actually the problem.
The healthier pattern is dual bookkeeping. Keep zones for machine files and broad strategy if that is how you farm. Keep named screening points for soil watchfulness. When a zone misbehaves on the yield map, ask which witness points inside that zone saw conductivity or pH move first. When a point misbehaves, ask whether the zone story still holds or whether you are looking at a sub-zone the polygon erased.
PAIRING LAYERS WITH SCREENING WITHOUT FALSE PRECISION
False precision is what happens when a smooth colour ramp implies certainty the underlying data did not earn. Yield at two tonnes per hectare resolution does not mean you know salinity at two tonnes resolution. NDVI at five-metre pixels does not mean pH is uniform inside each pixel. EM at twenty-metre swaths does not mean the same conductivity depth applies under every tree line.
Screening adds precision of a different sort: repeatability at a spot. If Wirraway West 03 reads consistently higher in conductivity than Wirraway West 01 under matched moisture, that is a real difference at those spots. Overlaying that difference on an EM map might confirm clay or salt trend — or might reveal that the EM layer missed a small saline seep because the survey was done in a dry month and the seep only expresses when the watertable rises.
Practical pairing steps:
Walk the map before you walk the paddock. Mark three to six places where layers disagree — high yield but low NDVI late season, high EM but average yield, stable yield but shifting NDVI year to year. Those disagreements are where soil stories hide.
Visit those places as named points, not as “somewhere in the red bit”. Record depth, moisture, crop or pasture, and the seven screening indicators if you use TerraProbe. You are not trying to paint the polygon. You are trying to explain whether soil at witness depth matches what the layer implied.
Hold season and management constant in your head. A yield layer from wheat after wheat tells a different story than wheat after fallow with a legume break. If you compare screening from a legume year to yield from a cereal year without noting the rotation, you will invent soil problems that are really rotation effects.
Use layers to prioritise laboratory sampling, not to replace it. When yield and NDVI both slump in a zone and conductivity screens climb at two witness points, that is a stronger brief for a salinity or sodicity package than either layer alone. When yield is fine but pH screens slide at a rise, the map will not warn you — the points will.
AUSTRALIAN CONTEXT: WHEN MAPS LIE KINDLY OR CRUELLY
In Mediterranean-climate dryland, one heroic season can make poor country look capable on a yield map, and the next drought can make good soil look hopeless. Multi-year yield stacks help, but they still average away the year when deep nitrogen or deep moisture saved a crop on shallow soil. Point screening that survives those seasons shows whether conductivity or pH actually changed, or whether the map only reflected rain.
In irrigation districts, uniformity is the fantasy every map sells. Centre pivots circle over zones that were never level, over sodic tails that only show up when application rates creep, and over bore water that shifts in salinity between summers. Pairing EC screens at fixed points with NDVI inside and outside the wetted circle catches edge effects that yield maps blur because the header still took grain from both.
In pastoral country, yield maps may not exist at all. NDVI or ground cover imagery still partitions the paddock. Screening at watering points, gilgai, and invasion fronts gives a soil record where “production” is wool or cattle weight, not tonnes per hectare. The same zone-versus-point logic applies: a bare patch on imagery is not automatically acidity; walk a named point before you commit lime.
If you advise several farms, say aloud that layers are property-specific. A conductivity colour scale that works on the Liverpool Plains does not transfer to a sand plain near Esperance without re-grounding. Witness points anchored to each property stop you from importing another farm’s zone map as gospel.
TERRATACTICS AS THE GLUE, NOT THE MAP
TerraTactics does not generate yield surfaces or satellite imagery. It holds screening history at named locations, with GPS assistance, moisture and depth notes, and charts that belong to the grower. That is intentional. The value is keeping point records stable while maps rotate through seasons and providers.
Beside a yield map on a laptop, keep the handoff honest: the map suggests where to look; the probe and routine decide what to measure; the Grower portal keeps the series when the map file is archived on a USB stick somewhere in the office. Optional TerraTactics Soil Analysis on Grower plans can summarise screening change — still insight, not a replacement for laboratory chemistry or variable-rate prescriptions.
For farm business records, exporting or simply maintaining a multi-year point history beside map screenshots in your own filing system is enough for many lenders and advisers. The credibility comes from named locations and consistent method, not from claiming the probe mapped the paddock.
COMMON MISTAKES WHEN LAYERS MEET POINTS
Treating one witness point as proof for an entire zone. Points justify investigation and targeted lab samples, not whole-paddock amendments from one reading.
Ignoring moisture when comparing screening to NDVI. A green image after rain can coincide with a conductivity spike that is mostly water, not salt. Note moisture every time.
Redrawing zones every year without updating witness points. If zones move but points stay put, you lose the thread of what changed.
Assuming EM depth matches probe depth. EM integrates through the profile; your standard 100 mm screen is not the same measurement. Use EM to choose where to point the probe, not to infer probe numbers without ground truthing.
Splicing yield from different headers, crops or moisture calibrations into one “master map” and then correlating to pH as if years were interchangeable. Fix the metadata or fix the habit.
Letting software auto-generate management zones without a paddock walk. The walk is where you discover the rock strip, the old burn, the strip of sodicity too narrow for the pixel.
A WORKFLOW FOR A TYPICAL SEASON
Pre-season: open last year’s yield and NDVI if you have them. Mark witness points for zones that underperformed or overperformed unexpectedly. Sample those points under your normal SOP before major inputs.
In-crop: if imagery dips in a zone that screened poorly at sowing, note whether it is uniform or patchy. Patchy often means machinery, pest or local drainage. Uniform often means nutrition or water. Screening mid-season is optional; when you do it, keep depth and moisture honest.
Post-harvest: yield map finalised (with whatever calibration you trust). Compare zone averages to witness point trends. Escalate to laboratory where point trends and maps agree something is wrong, or where they disagree and you need to know which lied.
Between seasons: do not delete points because they were “inconvenient”. Inconvenient points are usually the farm teaching you something.
FOR ADVISERS
When you sit with a grower who owns fancy maps but no point history, start with one paddock and three names. Show how a conductivity screen at a rise explains a chronic low-yield stripe better than another zone redraw. When you sit with a grower who owns probes but no maps, explain that screening without spatial context still beats memory, but layers help prioritise where to spend walking time.
Never sell TerraTactics as a yield-map replacement or a variable-rate engine. Sell it as the record that keeps maps accountable across staff changes and wet years.
BOTTOM LINE
Yield maps, NDVI and EM layers describe outcomes and proxies across space. Named point screening describes soil conditions at places you return to, with method you control. Zones help machines; points help truth. Pair them by letting layers suggest questions and letting repeatable screens answer whether soil at witness depth belongs in the story — then use the laboratory when money and risk demand proper chemistry.
False precision is the enemy. Honest pairing is the habit. Walk the disagreeing pixels, name the spots, screen with discipline, and let the paddock teach you what the colours only hinted at.
A screening reading is a starting point
TerraTactics readings help you target follow-up sampling. Confirm anything material with
accredited laboratory testing and a qualified local agronomist.
Published 04 July 2026 by David Gilbert
