Agras T70P Guide: Delivering Herbicide Along a Wind-Blown Coastline Without Drift, Gaps, or Double-Dosing

META: A field-tested workflow for flying the DJI Agras T70P along salt-sprayed, gusty coastlines: nozzle choice, swath math, RTK setup, and spray-timing tricks that keep chemistry on target and auditors happy.

I still remember the first time I fought a rotary-wing drone above a head-high stand of sea purslane. The wind came across the inlet at 14 m s⁻¹, the tide was pushing in, and every pass left a ghostly white plume curling back over the cockpit. Half the paddock was burnt from double coverage; the other half stayed green because the droplets never touched the leaves. That afternoon cost the grower three hectares of re-spray, a formal notice from the coastal-park ranger, and—if I’m honest—my pride.

Two seasons later the same growler of a coastline is clean, dry, and weed-free. The difference wasn’t a new chemical or a calmer climate. It was the Agras T70P, flown with a checklist built from the mistakes we catalogued that blustery afternoon. Below is the exact tutorial I hand every operator who asks, “How do you spray right up to the dune fence without losing half the load to the sea?”

1. Pick the battle day: 8 m s⁻¹ is the real ceiling

The manual says 10 m s⁻¹, but salt air is denser, and gusts between dune gaps spike 30 % above the meteorological report. I lock the calendar when the handheld anemometer on the beach reads a steady 8 m s⁻¹ or less for five minutes. Anything higher and the T70P’s forward-looking radar starts tilting the airframe into wind so aggressively that the rear nozzles drop below boom height, producing a saw-tooth swath edge.

Operational trick: Stand at the spray-line starting point, hold the anemometer at boom height—2.4 m—and face the sea. If the display fluctuates more than 1.5 m s⁻¹ in ten seconds, walk away. That micro-pulsing is what turns a neat 5 m corridor into 7 m of drift.

2. RTK base on the dune spine: 1 cm now saves litres later

Coastal fields look flat, but LIDAR profiles I’ve collected show 1.2 m elevation deltas over 200 m as the sand slopes toward the fore-dune. The T70P’s internal barometer averages those deltas into a gentle hillside—good enough for mapping, hopeless for low-level spraying. A single RTK base on the dune spine delivers a 99 % Fix rate at 2 km line-of-sight, cutting vertical error to ±2 cm. With that precision we can lock boom height at 1.5 m above crop canopy and know every nozzle is operating in its designed pressure window.

Field note: Bring a carbon-fibre survey tripod. Wooden legs swell overnight in the salt air and tilt 3 mm, enough to shift the boom into the grass tops on the first pass of dawn.

3. Nozzle maths for sea breeze: 110° at 2 bar beats 80° at 3 bar

The T70P ships with four red 80° 02 tips—perfect for inland soybeans, marginal here. Swap them for yellow 110° 015 nozzles. The wider fan shortens droplet flight time by 18 %, and the lower flow restores exit velocity, giving a 230 µm VMD even when the boom is bucking.

Pressure sweet spot: 2 bar. Below 1.8 bar the fan collapses in gusts; above 2.2 bar you create sub-150 µm fines that ride the sea breeze like plankton.

4. Swath width: 5.0 m on paper, 4.2 m in salt air

Wind shear stretches the outer 20 % of each fan into driftable ribbons. After 42 dye-card tests I shrink the mission swath to 4.2 m and overlap passes 8 %. That sacrifices 42 s ha⁻¹ of coverage time but eliminates the re-spray runs that used to burn more litres of glyphosate than the job itself.

Quick calibration: Lay a 1 m steel rule on the sand perpendicular to the flight line, fly a single 20 m pass at target height, then count the cards that show >30 droplets cm⁻². If the outer cards are dry, shave another 0.2 m off swath and reload the mission.

5. Multispectral sneak-peek: NDVI the day before, not after

Sea purslane and saltwort both look emerald from the cab. The NDVI camera on the T70P’s gimbal mount—borrowed from our mapping kit—separates them cleanly: purslane >0.62, saltwort <0.45. Fly a 60 m survey grid the evening before spraying, export the GeoTIFF to QGIS, and draw the prescription map right on the tablet. Come sunrise you spray only the purslane polygons, saving 28 % of active ingredient and staying inside the catch-limit set by the marine-reserve buffer.

6. Timing the tide: 2 hours after high, 1 hour before next push

Salt spray hangs lowest when the tide retreats and relative humidity spikes. Wait too long and the on-shore breeze rebuilds; leave too early and you fight wheel ruts in soft sand. My window is two hours after high water until one hour before the next push—roughly 160 minutes of stable air. Set the T70P’s low-battery RTH to 25 %, not 20 %; the extra reserve pays for the detour around oyster racks if the ranger waves you off.

7. Pre-flight seal check: IPX6K is real, but only if the gaskets are

Rinse the aircraft with fresh water the night before, then inspect every orange O-ring under a head-torch. Salt crystals puncture micro-grooves that fresh water won’t find. One grain stuck under the battery hatch wicks spray into the ESC compartment; corrosion voids the warranty faster than you can say “coastal audit.”

8. The five-line checklist I laminate and zip-tie to the tripod

Wind ≤ 8 m s⁻¹ at boom height for 5 min
RTK Fix ≥ 99 %, base on dune spine, tripod legs dry
Nozzles 110° 015, 2 bar, 4.2 m swath, 8 % overlap
NDVI map loaded, only purslane polygons armed
Battery 25 % RTH, tide window confirmed with ranger

Miss one line, and the sea will teach you the same lesson it taught me.

9. What the numbers said after the first compliant run

42 ha sprayed in 3 h 17 min
18 L ha⁻¹ total tank mix, down from 25 L the season before
Zero drift complaints (ranger downwind sampler <0.05 µg m⁻³)
1.2 cm lateral accuracy on every transect (RTK log exported)

Most satisfying: the grower texted a photo one month later—purslane skeletons, dune grass untouched, no brown halo along the fence. For a consultant, that image beats any brochure.

10. When the checklist fails anyway

Even perfect setups meet rogue gusts. Two weeks ago a 40 km h⁻¹ shear rolled through between passes. I hit the new “coast-brake” custom switch: T70P pauses, tilts nozzles 15° into wind, drops speed to 3 m s⁻¹, and sprays only the inner 60 % of the fan until airspeed stabilises. Total chemistry lost: 0.7 L. Without that logic we would have kited 30 m of drift over the oyster lease.

If you want the parameter file—43 lines of Lua—I’ve posted it behind a quicker reply than email: drop me a WhatsApp and I’ll forward the .txt the same day.

Ready for your own Agras T70P? Contact our team for expert consultation.