It watches the sky after legal light
Ducks migrate at night. We decode raw NEXRAD Level II volumes from the National Weather Service for the hours after sunset, when the sky over your region is empty of everything except weather and birds.
Duck migration forecast for ten states
Live NEXRAD radarEvery duck forecast on the market predicts. DuckCast predicts too, then checks itself against weather radar that watched the sky after dark. You get a score per region, the reasoning behind it, and an honest answer on whether last night backed it up.
The demo runs a real cold front through the season so you can see the whole product working. No account, nothing to install.
Radar check
Mass aloft, track and airspeed, read after sunset
Watch it work
How to read a regional score, what the eight factors behind it are, how the overnight radar check tells you whether birds actually moved, and how the freeze line warns you a push is coming before it reaches your state.
Recorded in the app, in demo mode, so the season is running. Nothing in it is a mockup.
1 min 50 sec · narrated
Why we built it this way
The easy version of this product is a handsome map, a five day forecast and some stock photography of a greenhead cupping in. That version is cheap to build and it tells you nothing you could not get on your phone already.
We built the other one. Every number on the screen traces back to something measured, and the two we work out ourselves instead of measuring say so, right there in the app, where it costs us something to admit it.
Raw radar volumes decoded in house, not a picture of somebody else's radar product
Five scans a night, per region
Real percent harvested for the crops that feed birds in your state
Weekly, in season
Hourly conditions at every region center and every upstream station
Hourly
Dry falls concentrate birds on what water is left. Wet falls scatter them
Weekly, published Thursdays
NOAA solar equations solved on our side, so legal light never depends on an API being up
Computed per date
Estimated from accumulated freezing degree days until the satellite ice feeds land. Labelled as an estimate in the app, because it is one
Rolling 12 to 14 days
Live crop harvest and the radar check are part of Pro. The free tier still gets a real score built on real weather, drought and season timing, using a typical harvest curve in place of this week's actual USDA figure.
The radar check
Weather radar has been read for bird migration for decades, almost entirely by ornithologists and conservation programs. The duck apps we looked at track collared birds by GPS or collect hunter reports. Neither is radar.
We have not found another duck hunting app that decodes raw NEXRAD volumes and hands the answer to a hunter over coffee the next morning. If you know of one, tell us and we will say so right here.
A forecast is a claim about tomorrow. It is worth a great deal more when something independent can tell you whether yesterday's claim held up. That is the job the radar does here.
Ducks migrate at night. We decode raw NEXRAD Level II volumes from the National Weather Service for the hours after sunset, when the sky over your region is empty of everything except weather and birds.
Bugs ride the wind. Birds fly through it. Subtract the wind at flight altitude from what the radar sees moving, and what is left is airspeed under something's own power. That is the number that matters.
Mass aloft, track direction and airspeed, read together, answer one question: did a migration happen over you last night, and was it headed your way or past you.
The honest limit
Radar is species-blind. It sees every bird in the air, not just ducks. Read the radar check as proof that a migration happened over your region, never as a duck count. We say this in the app too, in the same words, because a number you misunderstand is worse than no number.
The actual thing
Everything below is the real interface, rendered live from the demo scenario: early November, a strong Arctic front crossing prairie Canada, hard freeze and snow advancing behind it.
The morning hunt card
Tomorrow's Hunt · Thursday, Nov 5
Northwest North Dakota →Steady temps
Legal light 7:14 AMSunrise 7:44 AMSunset 5:29 PM🌘 Waning crescent · 20%
Best window: Legal light (7:14 AM) to 10:00 AM
No push wind forecast; first light is the best bet
Temperatures falling about 18°F over the next 48 hours
Times shown for the region center. Season dates and limits change yearly and by zone: verify with ND Game and Fish.
Regions, ranked
Big sheet-water country from Stanley toward Williston. Field mallard hunts build fast behind Saskatchewan fronts.
Souris River basin and pothole country around Minot. Classic early and mid-season traffic corridor.
The Devils Lake basin. Big water that holds mallards late, with heavy field feeds toward the Red River Valley.
Tap any region for the full breakdown: every factor, the upstream stations, the freeze line and last night's radar read.
How a score is built
Every regional score is eight named factors, each rated zero to ten and combined at a fixed weight. Open any region and you can see which factor moved the number and by how much.
We publish the weights because a forecast you cannot argue with is a forecast you cannot trust. If the model says a nine and your marsh is empty, you should be able to see exactly which input oversold it.
See it on a real frontHunter reports carry 12% of every score, which is why filing one is free at every tier. Charging for the right to contribute would thin the forecast it pays for.
Coverage
Each state gets its own regions, its own upstream weather stations, its own radar sites, and season and crop curves fitted to it. That work does not copy and paste, so we would rather cover ten states honestly than twenty badly. More are coming, one properly built state at a time.
Conditions across southern Saskatchewan and Manitoba usually predict arrivals here 24 to 72 hours out.
Conditions across Missouri, southern Illinois and west Tennessee usually predict arrivals here 24 to 72 hours out.
Missouri fills from two directions at once: down the Illinois River and the upper Mississippi pools in the east, and down the Missouri River through western Iowa in the west.
South Dakota fills from one state and one state only. Everything arrives down the James and the Missouri out of North Dakota, which means the front that closes the season up there is the same front that starts the best week down here.
Nebraska fills from one state, and it is one DuckCast already scores. Birds come down the Missouri and the James out of South Dakota, which means the same front can be followed from prairie Canada through North Dakota and South Dakota to the moment it drops birds into the Rainwater Basin.
Kansas fills from one state, and it is one DuckCast already scores. That makes four in a row: a front can be followed from prairie Canada through North Dakota, South Dakota and Nebraska to the night it drops mallards onto Cheyenne Bottoms. The eight Nebraska stations sit at two latitudes because the same freeze is two or three days out from central Nebraska and about half a day out from the state line.
Mississippi fills from the Arkansas Delta and west Tennessee, straight down the river. Arkansas is a state DuckCast already scores, and so is Missouri above it, which makes this the bottom of a second chain: Missouri to Arkansas to Mississippi. When the Grand Prairie freezes, this is where those birds land.
Conditions across south Arkansas, the Mississippi Delta and northeast Texas usually predict arrivals here 24 to 72 hours out.
Conditions across Kansas and southeast Colorado usually predict arrivals here 24 to 72 hours out.
Conditions across Oklahoma and eastern New Mexico usually predict arrivals here 24 to 72 hours out.
We would rather hear it from hunters than guess from a map. Tap every state you hunt. The vote takes no email and no account, and stores nothing about you.
One tap. This vote takes no email and stores nothing about you. Pick as many as you hunt.
Somewhere else, or want to tell us what the local read actually turns on, the reservoir that holds birds until it locks, the river that never does? Send it over. That is the sort of thing you cannot look up.
Free guides
The forecast covers ten states. These guides cover all of them: when each state's season opens, when the birds actually move, and what turns the migration on. Every date traces to the agency that set it.
Every US state's duck season in one place: verified 2026-27 dates where agencies have published them, September teal openers, zone structures, and the official regulations link for all 50 states.
Read the guide →
Blue-winged teal leave in September, gadwall and wigeon ride the first fronts, mallards follow the freeze line into December and green-winged teal outstay them. What actually triggers the duck migration, how far ducks go, when they come back north, and when the push reaches each flyway.
Read the guide →
Blue-winged teal migrate before any regular duck opener, which is the whole reason the early September teal season exists. 2026 dates, limits, and how to hunt the earliest ducks of the year.
Read the guide →
Your spot stays yours
The fastest way to kill a hunting app is to turn it into a map of where everyone is hunting. DuckCast forecasts conditions across a region. It never was, and never will be, a spot finder.
On your phone
The same scoring engine that runs this site, in an app designed for where you actually read it: dark, cold, early, and usually somewhere the signal is not.
Requires iOS 17 or later. Subscriptions are handled by the App Store.
Amber on warm black, and nothing on screen is allowed to go blue-white. You can check the wind at 5 AM without resetting the night vision you spent an hour earning.
The last good forecast is kept on the phone and shown with its age, so a blind with one bar gives you a two hour old score instead of a spinner.
High contrast for sun off snow and open water, where a handsome dark interface becomes a mirror you cannot read.
Common questions
Most of a DuckCast score is weather: what the air is about to do, and what that usually does to birds. The radar check is the one input that measures the birds themselves.
We decode raw NEXRAD Level II radar volumes from the National Weather Service and read what was actually in the air over your region after sunset last night: how much bird mass lifted off, which way it tracked, and how fast it flew under its own power once the wind at flight altitude is subtracted out. Bugs drift with the wind. Birds fly through it.
Hunter reports are the other set of real eyes on the birds, and they carry 12% of every score, but they depend on somebody having been out there. The radar watches every night, across the whole region, whether anyone hunted or not. One honest limit, and we say it in the app too: radar sees every bird in the air, not just ducks.
Read it as proof that a migration happened, not as a duck count.
Two reasons. Privacy first: exact hunting locations are never collected or shown, every report is generalized to the region level, so nobody burns your spot. Biology second: migrating flocks ride weather systems that span whole counties. A regional score tells you when birds are moving through your area. Finding the X is still your job.
It keeps running. The freeze line tracks all winter, and the radar still reads the sky every night, which is how a real movement in the dead of summer turns up at all. What changes is the scores, and they stay honest rather than manufactured: outside the migration window there is no migration, so the number is low and the app says why.
Two things we want and have not built yet are season recaps for your saved regions, and a model trained on past seasons to sharpen the next one. Today's engine is a published set of weights, tuned by hand. We would rather tell you that now than have you find out in February.
No, never. Hunter reports exist for one purpose: improving the regional forecasts for the people who file them. We do not sell reports, contact info, or location data, and exact locations are never collected in the first place.
Yes, for the states DuckCast covers. The live map colors every one of the 90 regions across North Dakota, Arkansas, Missouri, South Dakota, Nebraska, Kansas, Mississippi, Louisiana, Oklahoma and Texas by its current migration score, rebuilt through the day from live weather, the freeze line across the country upstream, crop progress and hunter reports.
Each morning the radar check marks whether birds actually moved over a region after dark, so it is a forecast map with a nightly observation laid on top rather than a map of sightings. It covers ten states today, not the whole country, and it says so instead of painting states it does not model.
Every covered state has its own page listing the regions DuckCast scores, and every region has a page describing what the forecast watches there. The report itself is the live score: a 0 to 10 number per region, the eight factors behind it, and last night's radar read. Hunter reports filed in the app are generalized to the region and feed the same number.
Your state wildlife agency publishes the official regulations and, in most states, in-season survey counts; DuckCast links to the agency from every state page and never stands in for it.
Yes. DuckCast is a duck hunting app for iPhone and the web that forecasts the migration for ten states, region by region, and checks itself against NEXRAD radar every night. It is not a land-ownership or public-land map and it does not sell gear or book guides. The web version is free to start with no account.
The iPhone app is on the App Store for iOS 17 or later, and Pro adds the live radar check, live crop data, the live freeze line and reading other hunters' reports.
Take ten minutes with the demo. It runs an Arctic front through a real season so you can watch the scores, the freeze line and the radar check all move together, which is the only way to judge whether this is worth your money.
Built by hunters in Georgia, for hunters in North Dakota, Arkansas, Missouri, South Dakota, Nebraska, Kansas, Mississippi, Louisiana, Oklahoma and Texas. More states are coming, and the vote above is how we decide the order.