/* ============================================================
EMI Soil Scanning — Electrical Conductivity at three depths.
The foundation every other layer sits on.
============================================================ */
(function () {
window.EMIPage = function EMIPage({ setRoute }) {
return (
<>
EMI scanning fills the gap your 50 × 50 m chemical grid leaves behind — mapping soil Electrical Conductivity at three depths (25, 50 and 90 cm) at a practical 5 × 5 m resolution. The foundation every other layer sits on, and the first layer your RevToolbox loads.
Where standard chemical grids only sample every 50×50 m, EMI scanning fills in the gaps — mapping soil Electrical Conductivity at three depths (25, 50 and 90 cm) at a practical 5×5 m resolution. The foundation every other layer sits on, and the first layer your RevToolbox loads.
EC at 25, 50 and 90 cm — root, sub-root, drainage zones
Where soil EC and canopy NDVI agree, you've found a real zone.
EC reveals what the soil can do — texture, water-holding, drainage. NDVI reveals what the canopy actually did. Stack them and the pattern that shows up in both layers is your management zone. The pattern that shows up in only one is your sampling priority.
NDVICanopy vigour
2024-11-15 → 2025-02-030.59 → 0.74 index
VS
EC 50Soil conductivity · 50 cm
5 × 5 m resolution3.04 → 9.73 mS/m
ALow NDVI + low EC
Sand or shallow soils — chronically low vigour. Drainage and rooting volume, not fertiliser, are the limits here.
BHigh NDVI + mid EC
The block's good soil. Match inputs to expected yield; this is the baseline every other zone is judged against.
CEC and NDVI disagree
The soil says one thing, the canopy says another. Sample first — irrigation, salinity, or a buried chemistry issue is hiding here.
{/* ============== EMI · combo examples → lead to zonal management ============== */}
Examples from working orchards
Put two layers side by side. The block stops looking uniform.
Each block below is scanned with EMI and seen by the satellite. Read the soil against what the canopy and the crop actually did, and the same truth shows up every time — the field is not one thing, and managing it as if it were leaves yield and money on the table.
01High EC drives strong vigour — and weak fruit setOne block · the soil layer, read twice
01
Where the soil EC runs highest, the canopy grows strongest.
EC, left · 11.35–26.6 mS/m · NDVI, right · 0.6–0.82. The high-EC core holds more water and more nutrients — and the canopy answers with the darkest green on the block.
but
02
…and the very same core gives up its fruit.
EC, left · 11.35–26.6 mS/m · Yield, right · 1.82–12.1 · 2023-11-10. The core runs red — that vigour went into wood and leaf and fruit set fell away. The lighter, low-EC edges grew less but set and carried more fruit.
One block. Two opposite needs. No single blanket rate can satisfy both.
02A different block — read the sub-soil and you can see the drainageCanopy NDVI · vs · deep EC
↳
The strongest canopy sits over the better-draining sub-soil.
NDVI, left · 0.35–0.52 · EC, right · 16.96–34.84 mS/m. Where the deep EC reads high, water and nutrients are held where the roots can reach them — so the canopy is strongest right there. Read the conductive layer and the drain path is visible before you ever cut a trench.
Every one of these blocks asks for different inputs in different corners. That is the whole case for zonal management — so the only question left is where do the samples go?
{/* ============== EMI · From overlay to management zones ============== */}
From overlay to zones
Stop sampling on a grid. Sample by zone.
Once EC and NDVI are stacked, the platform draws the management zones for you — groups of trees that behave the same way. Sample inside the zone, not on a grid that ignores it: a grid point can land right on the border between two soils and represent neither, while several sub-samples composited inside a zone describe its chemistry honestly. The payoff comes at the spreader — it applies the right amount right up to the edge of the soil boundary, instead of feathering an interpolated 50×50 m average across a line it can't see.
5×5 m EC pixels reconciled with 10 m NDVI
3–7 zones per block, automatically delineated
6–12 targeted samples instead of a 50×50 m grid
1 prescription file per zone, per product
EC + NDVI overlaySample pins placed by zone, not grid · 8 farms, 301 ha
{/* ============== EMI · Zoning technique (real case study walkthrough) ============== */}
The soil truth. Conductivity at 5 × 5 m resolution shows where the texture, water-holding and salinity actually sit — independent of what's planted on top.
02
Two NDVI snapshots
NDVI · period A
NDVI · period B
The canopy truth, at two different points in the season. Comparing snapshots filters out one-off events and exposes the patterns the trees show every year.
03
Zones where they agree
Zones · red boundary
Where soil EC and canopy NDVI agree, you've found a real management zone. The algorithm draws them; the agronomist confirms them.
04
One composite sample per zone
Sample points
Sub-samples within a zone are composited into one — capturing that zone's soil chemistry honestly, without the noise (or cost) of grid-based sampling.
Scales across the operation
One recipe, every block.
The same four-step technique runs on a single 3-ha block or 300 ha of mixed cultivars and ages. Every block gets its own zones and its own sample plan — built from the data, not the grid.
301 ha8 farms · case study
6–12samples per block · not 50×50 m
1×composite · per zone
{/* ============== EMI · ROI teaser — case study pull ============== */}
What this unlocks
Zone-by-zone prescriptions. R3,344/ha avoided. Every year.
EMI + NDVI is the foundation. The product is the prescription that comes out the other end. Across eight farms and 301 hectares, the same technique you just saw avoided R5.0 million in misallocated amendments — a 9.1× return on the scan and sampling spend.
9.1×
Return per R1 invested
EMI + zone sampling, 5 years
R3,344
Avoided per hectare / year
vs. blanket application
66%
Blanket spend wasted
K₂SO₄ · MgO · Gypsum
Example · Block 3 · 7 zones · 5.4 ha
Seven zones, seven prescriptions.
Zone D is severely K-deficient (1.1%) and needs 5,481 kg/ha K₂SO₄. Zone G already has adequate potassium (4.8%) and needs none — only 320 kg/ha gypsum to correct sodium. A blanket average would spread 2,852 kg/ha K₂SO₄ across all seven zones — under-dosing the deficient zones by up to 48% while Zone G receives nearly three tonnes of product per hectare it can't use.
One block, seven prescriptions. No composite sample can capture this.
