See which layers Phase Spatial can use, look at them over the paddock, and import more yield maps, NDVI dates and files for one paddock or all of them. Phase Spatial comes with Phase Ultra and the free trial.
Before you start
Open More → Paddocks & crops → Spatial, then tap the paddock on the farm map (or choose it in the picker at the top). Only paddocks with a mapped boundary are there.
Import layers is on the Layers step and in Actions (on a phone, Import in the Layers tab). John Deere imports need John Deere connected.
What Spatial can use
Yield maps with at least three usable cells. Its year follows the harvest season, not the upload date.
Soil results, one layer per sampling year, from sites with a location: pH, Colwell P and K, organic carbon, sulfur, salinity (EC), CEC, aluminium, PBI and more, with depths where the lab gave them. See Use soil data on the paddock, in Compare and in Spatial.
Satellite NDVI, one layer per date you pull.
Uploaded files: yield maps off the header, EM38 and EC surveys, NDVI point files, lab results, prescription files, and zone maps from a contractor, an agronomist or your current farm software. A soil-type or management-zone map keeps its zone names (Sand, Loam, Mallee), each in its own colour on every paddock.
Elevation, and job maps from John Deere and CNH FieldOps machines (below).
Look at your layers
On the Layers step, switch a layer on to see it over the paddock. One layer (or saved zone map) shows at a time: switching one on switches the other off, keeping its colours for next time. To use several layers together, build zones from them.
On a phone each layer has an on and off switch, with Rain, Frost and Wind under See it in 3D below them.
The colour button names the layer's colour scale; press it for the next. Layers open in Viridis (dark to light) unless you'd coloured them already.
A soil layer has Measurement and Depth choices.
Display holds Opacity, Colour bands and Smoothing. These change only how the layer is drawn, never the readings or a zone map.
Delete this layer, also under Display, shows on yield maps and layers you imported. It removes the layer on every device, and a zone map built on it is rebuilt without it. For a yield map, the season's yield figures stay. To get a layer back, import it again.
Saved zone maps are under Zone maps, below the layers (on a phone, the Zone maps tab).
Yield map and NDVI together. Only one shows at a time, but Read lists every layer's value at a spot (see Read a paddock, see it in 3D and ask Phase Agent in Spatial), and building zones from both blends them (see Build and edit a zone map).
Upload a yield map, survey or zone file
Select Import layers → Upload → Browse.
Choose a shapefile (a .zip, or the .shp and .dbf together), a GeoJSON or JSON file, or a CSV or TXT with latitude and longitude columns. The file then goes in three steps, What's in it, How it looks and Check and add, with Back and Continue. Nothing is saved until the last one.
On What's in it, check What kind of layer is this?, such as Yield, EM / soil conductivity or Soil type / zones.
Pick the column under Which column holds the measurement? (for a zone map, Which column tells the areas apart?). Each column is a card that says what it holds, such as "SOILTYPE: Loam, Mallee, Sand and 1 more", and the one Phase suggests is marked Suggested. A yield export also carries speed, moisture and swath width.
On How it looks, check the file on the satellite map, in its colours, over the paddocks it lands on. For a yield map, choose the Season and Crop. For anything else, check the Layer name, Unit, Colours and Which readings make the top zone?
On Check and add, check the name, what it holds and which paddocks it goes to. Select Save yield map to paddock for a yield map, or Add layer (Add to 3 paddocks for a file over three) for anything else.
A yield map goes on the paddock on screen, on its yield record for that season, so Yield Centre shows it too; tonnes already entered are never changed. It is cleaned first (each run's start and end, turns and part-width passes are set aside), and the message says how many readings went. If its average is more than about 1.5% off the paddock's yield, the message offers Match the map, then Undo. See View and use yield maps. Other uploads are saved to the paddock for every device.
Protein. A yield file with a protein sensor's column (such as "Protein" or "CrdPrPrcnt") also makes a Protein layer in %, named such as "Protein · 2026 Wheat"; Check and add says so. It never changes the protein on the yield record. Uploading the same file again for the season replaces it. A column of zeros makes no layer.
Zone names. A map whose zones are named comes in with them: each name gets its own colour, the legend lists them, and Read shows "Sand", not a number. If the column you pick is the same on every zone, Phase warns the map will be one colour: pick the one that differs.
Contractors' files. A nutrient map with low and high values per band is read at the middle of each band, and a prescription shapefile by its rate column. A zip of several shapefiles reads the first and names the rest. -9999 "no data" values are left out.
A file covering several paddocks. Upload it from any one paddock. Phase says where it will go first, such as "This file covers 3 paddocks: North 1 (4 zones), North 2 (3 zones) and Creek Flat (2 zones)." Each paddock gets its own layer and colour scale; anything off your mapped paddocks is left out and counted. A file on no mapped paddock goes on the paddock on screen, with a note to check it lines up.
Pull NDVI dates from the satellite
Select Import layers → Satellite, with the paddock's own name chosen at the top (marked This paddock). Phase lists the Sentinel-2 passes over the paddock in the last year by month, with cloud cover. Over 40% is marked "likely patchy"; a date you have says In layers.
Tick every date you want.
Select Pull 3 dates as layers (the count changes).
Each date becomes its own NDVI layer. Cloud is left unmapped, not read as low vigour.
Import a yield map from John Deere
Select Import layers → John Deere, with the paddock's own name chosen at the top.
Under Operations Center paddock, choose the matching field.
Choose a harvest. Look back further than 5 seasons shows older ones.
Most maps take under a minute, a big paddock up to 30. You needn't wait: Phase keeps asking for up to six hours and adds the map when it's ready. Waiting maps show under Still being built with Check now; a failed one says why, with Try again.
Import for every paddock at once
On the John Deere and Satellite tabs, choose All N paddocks instead of the paddock, then pick the Season, so last year's maps can come in from this year's screen. Yield Centre → Yield → Map → John Deere maps fetches them too, but files every map under the one season you choose there.
John Deere:
Check Harvest from and Harvest to (widen to the end of January for canola stripped after New Year). A window over several harvests files each map under its own crop year's season, making one such as "2025" if the farm has none.
Select Look for harvest maps. Nothing is fetched yet. Each paddock is matched to Operations Center by name (or by its linked field where the boundary came from John Deere) and marked Ready, In Phase (already has a map for the season), Not matched, No harvest or Failed.
Select Fetch N yield maps for N paddocks. Each map is saved to its paddock as it finishes, usually within minutes. You can close the screen. When the maps span more than one harvest, the message says which seasons they went to.
Satellite: tick the dates and select Pull N dates over N paddocks. Each paddock gets each date as its own NDVI layer, skipping dates it has.
Elevation
Phase uses the best ground height it has: your own survey or RTK file, then your John Deere machines' heights, then OpenTopography's 30 m satellite model. Spatial, the rain, frost and wind runs, and Yield Centre's 3D view all use it.
Satellite model. A paddock with no elevation lists Elevation marked OpenTopography: "Ground height, 30 m. Switch on to fetch it". Switching it on, or adding it to a zone map, fetches the ground for every paddock on the farm, and it then stays for everyone. Fetching needs permission to edit the farm.
John Deere machines. On Phase Ultra with John Deere connected, the ground height logged on a job becomes the paddock's Elevation · John Deere layer, from the job covering most of the paddock. With RTK it's accurate to a few centimetres; without, it can sit a little high or low but its rises and hollows are right.
Machine job maps
On Phase Ultra, each John Deere job Phase has read, and each CNH FieldOps harvest, leaves its maps on the Layers step, newest first, one row per pass. They sit under John Deere jobs (CNH jobs, or Machine jobs when you have both): Applied (one per product), Target on a variable-rate job, Yield on a harvest, Diesel and Speed, and Coverage on a CNH harvest. A John Deere harvest from a combine with a protein sensor also has Protein, each square the protein of the grain off it. Buttons such as Spray and Harvest show one kind of work at a time.
Every job map, and every yield map from a harvest confirmed in Reconcile, is marked John Deere or CNH FieldOps. On the Zones step the job maps sit together under Machine job maps, ready to build from. See John Deere harvest jobs, diesel and job maps.
If something doesn't look right
An import doesn't work. Email the file to support@phase.ag and we can help bring it in.
My agronomist sent a PDF of VR or zone maps. Phase can't read maps from a PDF or a picture. Ask for the shapefile or Rx files and upload those.
A KML, KMZ, GeoPackage, GeoTIFF, spreadsheet or ISOXML file is refused. Export a shapefile, a GeoJSON, or a CSV with latitude and longitude instead. Phase says which.
"Those coordinates are eastings and northings". The file is in MGA or UTM without its .prj. Include the .prj, or re-export it in WGS84.
"That zip has no shapefile (.shp) in it". It holds something else, such as an ISOXML TASKDATA folder; Phase names what. Ask for a zipped shapefile.
"Phase needs … .dbf as well". Choose the .shp and .dbf together, or the .zip.
Soil results "can't be mapped yet". They have no GPS. Select Link results to zones: soil-test zones link through Soil Centre, not Upload.
No passes found. Check the paddock boundary.
A paddock shows Not matched. Give it the same name in Operations Center, then select Look again.
A John Deere map wasn't saved. One of only a few hundred readings (the header crossing the paddock), or over 20 t/ha for a grain crop, is refused with the reason. "That harvest has no yield readings in it" usually means it was cut for hay or the monitor was off.