Variable-rate nitrogen from soil, landscape water, this season's weather and your N plan — compared with KdN V3.3.1
BoundaryTerrainSoil testYield
1Field boundary
Shapefile (.zip, or .shp + .dbf + .prj together), GeoJSON or KML.
2Terrain analysis
The buffer pulls in ground outside the boundary so water running onto the field is counted.
3Soil test
Point shapefile (.zip), CSV with latitude/longitude, GeoJSON or KML. CSV template
4Yield
Points only: averages raw yield points over this radius before the Rx is built (0 = off).
Points (yield map, CSV), polygons (SMS MRY grid, yield zones) as shapefile .zip or GeoJSON, or a GeoTIFF in lat/long. Values in bu/ac, or tick “relative” above.
5Zone & field
Zone is suggested from the field location when the boundary loads. Suggest again
lb N/ac. Other N = starter, MAP/DAP. Soil nitrate is credited by zone (full in B, C, D; half in A and E).
Observed weather since Oct 1, then 30+ past seasons for the rest of the season. Losses are compared with typical ground, a typical year and spring preplant NH3 — the conditions V3.3.1 was tuned under.
7Economics & output
Low-zone guard: N demand never drops below this % of the field-average yield. MRTN pull shifts the whole field toward the university rate and keeps the spatial pattern.
No terrain built: every cell is treated as typical ground (W = 0).
Season-N · field-testing build. Soil and yield are interpolated by inverse distance; terrain is built from USGS 3DEP lidar or yield-monitor elevation. Weather, the season simulation and the rate model run on AvlingAI servers. Loss rates, zone coefficients and the price adjustment are starting values being calibrated with strip trials.
SeasoN
Sign in with your AvlingAI account. We'll email you a link and a 6-digit code — no password.