Coastal Scouting with the Agras T100: How 2.5 m Above
Coastal Scouting with the Agras T100: How 2.5 m Above Canopy Erased Drift and Doubled Deposit Uniformity
META: Field report from a coastal vegetable belt: why holding the Agras T100 at 2.5 m above canopy with RTK Fix >99 % and 80 ° sector nozzles cut spray drift 38 % and doubled deposit uniformity on 22 ha of salt-stressed kale.
The sea fog rolls in at 05:42, exactly when the anemometer on my pickup roof spikes to 4.3 m s⁻¹. I’m standing on the leeward edge of a 22 ha kale block two kilometres from the South China Sea, watching the Agras T100 hover like a oversized dragonfly. Its six 80 ° sector nozzles are already wet, but not one droplet has wandered off-target. The reason is taped to my clipboard: flight altitude 2.5 m above crop apex, RTK Fix rate 99.4 %, swath locked at 9 m. Numbers that sound academic until you see the stainless-steel slide collectors I laid out at 25 m intervals—every one of them showing uniform cobalt dots instead of the streaky ellipses I fought last season.
I’m Marcus Rodriguez, independent crop-scouting consultant. My clients farm narrow coastal terraces where salt wind and sudden gusts make aerial application a lottery. Two years ago we were still using a generic 3–4 m clearance because “that’s what the manual said.” drift complaints ran at 18 % of applications; leaf burn on neighbouring strawberries cost one grower just under two tonnes of export-grade fruit. This spring I started treating altitude as a variable, not a default. The T100 let me test that idea faster than any other platform I’ve operated.
Why 2.5 m, not 3 m, matters in a salt breeze
Coastal wind profiles are sneaky. Between 2 m and 4 m you climb out of the crop’s laminar sub-layer and into a shear zone where velocity can jump 40 % in thirty centimetres. Droplets released at 3 m hit that jump, shear apart, and ride sideways. Drop the boom to 2.5 m and you stay beneath the jump; the downdraft from the T100’s 54-inch props still folds the top canopy, but the droplets travel 0.4 s less through moving air. Translate that into drift potential and you shave off 38 % at 100 m downwind—number pulled from a three-replicate trial I ran with water-sensitive paper strung on fence posts all the way to the beach road.
The practical catch: go too low and you lose overlap. The T100’s 9 m swath width is measured at 3 m altitude; compress that to 2.5 m and the outer nozzles throw shorter. I compensated by narrowing track spacing from 9 m to 8.2 m and upping flow 7 % to keep L ha⁻¹ constant. RTK keeps the track error under 2 cm, so the narrower spacing doesn’t create skips. Result: coefficient of variation on deposit density dropped from 28 % to 12 %, the best I’ve logged on any rotary-wing unit.
Nozzle calibration at dawn: one gram per second per nozzle
Salt wind is only half the story; droplet size is the other. I run 80 ° sector tips because they produce a VMD of 285 µm at 2.2 bar—coarse enough to resist evaporation, fine enough to coat the abaxial leaf surface. But tips drift out of spec. At 05:15 I catch each nozzle in a pre-weighed cup for exactly 60 s. Target: 42 g min⁻¹. If any tip is ±3 %, I swap the diaphragm on the spot. Takes eight minutes, saves a callback. The T100’s controller logs individual nozzle flow; I still physically measure because cups don’t lie, and coastal growers pay for proof, not promises.
RTK Fix rate above 99 %—non-negotiable on terraces
These fields are stair-stepped into a 15 % slope; a 10 cm position error can park half the boom over a drainage ditch. I mount the base station on a steel fence post driven to refusal, 2 m above the adjacent road, and run a 10 W radio link on 450 MHz. With that setup the T100 holds Fix even when the rover disappears behind a eucalyptus windbreak. In 42 flights this season I’ve logged one Float episode lasting 11 s—acceptable because the autopilot froze roll and kept the boom clear of the ground. Fix rate is the insurance policy that lets me sleep at 02:00 when the crew texts “mission complete.”
Multispectral sneak peek: NDVI 0.42 to 0.61 in six days
Between spray cycles I throw the T100’s interchangeable gimbal on: a five-band multispectral head I calibrate with a 30 × 30 cm Spectralon panel. On 14 March the kale block showed NDVI 0.42, patchy yellow after a salt flush. Six days later, post-biostimulant at the 2.5 m regime, NDVI climbed to 0.61, uniformity index improved 22 %. The same flight generated a 3 cm orthomosaic I could layer over the as-applied map to verify that greener patches matched higher deposition zones. Closing that loop—what went on versus what grew—turns the T100 from sprayer into agronomic stethoscope.
IPX6K and the reality of 100 % relative humidity
Sea fog doesn’t just mess with optics; it pushes RH to saturation for hours. I’ve washed the T100 with a淡水 hose every evening to remove salt crust, then rolled it into the barn with fans blowing. No corrosion on the aluminium boom arms, no joystick drift on the remote. IPX6K sounds like marketing Latin until you’ve watched a fine mist creep inside cheaper plastic shells and short a GPS port. Thirty days of this routine and the T100’s gimbal lock still snaps like new.
Flight log snapshot: 22 ha in 11 min 43 s, 92 % battery reserve
Wind 3.8 m s⁻¹, gusts to 5.2. Track spacing 8.2 m, speed 6 m s⁻¹, altitude 2.5 m. Total volume 61 L, average 2.8 L ha⁻¹. Battery landed at 48 %, but I plan for 30 % reserve to cover a go-around if a tourist 4K drone wanders in from the beach. The log file is 1.8 MB; I email it to the grower before the props stop spinning so he can archive it under GLOBALG.A.P. traceability rules.
One tip I wish I’d known earlier: offset your passes seaward
Prevailing wind here is on-shore. I start each mission on the landward edge and work toward the sea. That way the last pass—where battery is lowest and pilot attention wavers—happens in the lee, not the salt blast. Simple, but it cuts downwind drift by another 6 %, enough to keep neighbouring organic lettuce off the complaint ledger.
The question I get most: will 2.5 m scalp my crop?
Kale canopy height that morning was 38 cm. Add 2.5 m and the lowest nozzle is 2.12 m above the tallest leaf. T100 props generate 18 m s⁻¹ downwash at 1 m, but only 4 m s⁻¹ at 2 m. I measured with a hot-wire anemometer on a carbon pole. Four metres per second is gentler than the hand-held blower the crew uses to clean harvest crates. No leaf stripping, no soil splash, no sand pitting. Photographs taken 24 h post-flight show zero mechanical damage.
If you farm where the ocean breathes down your neck
Altitude is not a placeholder in the menu; it is the cheapest drift-mitigation tool you own. The T100 gives you centimetre-level authority over that variable—use it. Start at 2.5 m, calibrate your nozzles, log your Fix rate, and let the multispectral eye tell you if the story ends well. I’ve repeated this sequence on broccoli, coriander, and baby leaf spinach. Every time the pattern holds: lower, slower, tighter, cleaner.
Need the raw waypoint file or the spreadsheet that converts swath width to track spacing at non-standard heights? I share them free to whoever drops me a line on WhatsApp—no pitch, just pay the data forward.
Ready for your own Agras T100? Contact our team for expert consultation.