WHAT'S HAPPENING NOW
—— ft above typical
Finding the top signal on the water…
Live river level, flow, historical records and forecast for Brazos River near Hempstead in TX, United States, built from official government gauge data.
The Brazos River near Hempstead sits in the Lower Brazos-Little Brazos watershed in Texas.
The river usually runs highest in May at 15.3 ft (4.66 m), then lowest in August at 11.21 ft (3.42 m). That typical annual swing is 4.1 ft (1.25 m), which is the range a dock, pump intake, or low bank setup has to handle through the year.
For people watching water levels, this is a river that spends part of the year on a lower summer track and then builds into a higher spring stage. In a flash-flood setting, rainfall can change the level quickly, so long quiet stretches can end fast after a storm.
The record high is 53.83 ft (16.41 m), set in May 2016. The record low is 7.76 ft (2.37 m), set in November 2014.
Those two marks show the spread this station has seen over its record. The high tells you what the river can reach in an extreme event. The low tells you how far stage can fall during a dry stretch. For anyone with gear near the bank, those are the two edges that matter most.
Fastest 24-hour rise is not listed in the facts here, so I won’t guess at warning time from this station.
The hero number on this page is today’s reading set against the median for this calendar day, not raw elevation by itself. That comparison is more useful because a stage that looks ordinary in one month may be high or low for another month.
This station is a USGS site, and it has more than 50 years on record. The upstream drainage area at the station is 43,880 sq mi (113,649 km2).
Readings come from official government gauges and are rolled into daily records: the daily mean, high and low for every day on record. Each calendar day gets its own baseline from the full history, so the level you see is compared against what this specific date usually looks like rather than an annual average.
Real-time water levels, forecasts, and alerts for lakes and rivers across Canada, the US and the UK. Built on official government data.
Nibi Metrics tracks more than 13,000 gauged lakes and rivers across Canada, the United States and the United Kingdom, reading directly from official federal monitoring networks: the Water Survey of Canada, the United States Geological Survey, Quebec's provincial hydrometric network, and the Environment Agency for England. Every number on the platform is measured, not modelled - taken from a government gauge on the water itself and compared against that location's own historical record for this exact date.
The level right now, in feet or metres, updated as the source agency publishes new readings - for many stations that is every five to fifteen minutes. How today compares to typical: every location carries its own typical-for-this-date band built from its historical record, so a reading is never just a number - it is above, below, or near what this water usually does on this calendar day. Which way it is moving: a measured RISING, FALLING, or STEADY verdict computed from the gauge's own recent readings, with the rate of change - the difference between a flood ending and one beginning. What comes next: level forecasts up to 72 hours ahead, driven by the gauge's recent behaviour, upstream water, and quantitative precipitation forecasts, alongside live radar, lightning, and weather for the surrounding region.
Set a threshold on any location - a level, a distance above typical, or a rate of climb - and Nibi emails or notifies you the moment the gauge crosses it. Cottagers use it for spring freshet, anglers for river conditions, marinas and municipalities for operations. The scanning runs around the clock against the same measured data shown on every page.
Nibi means water in Anishinaabemowin. The platform began at a family cottage on Lake Muskoka, Ontario, where the spring melt floods the shoreline most years and the only forecast was standing on the dock, looking, and guessing. It grew into an independent, member-funded platform run by Muskoka Research Inc. - no investors, no data resale, and a privacy policy shorter than most cookie banners. The core data is free for everyone, forever; paid plans add automation like multi-location alerts, deep history, and reports.
Nibi spans the continent: every Canadian province and territory with active federal gauges - from Vancouver Island rivers to the Muskoka lakes, prairie reservoirs, and Quebec's hydrometric network - and the United States from Maine to the Texas Hill Country, the Ohio Valley, and the Pacific Northwest. Coverage grows where the water news is: when a basin floods, its gauges join the platform with their full historical records attached. Many stations report every five to fifteen minutes; most report at least hourly; every page states exactly when its gauge last spoke.
The core promise is simple: the water data is free, for everyone, forever. Live levels, maps, radar, and recent history never sit behind a paywall. Paid plans fund the platform by adding automation - multi-location alerts, decades of history, longer forecasts, and PDF reports - so the people who need more carry the cost for everyone who just needs to check the lake before driving up. There are no investors and no advertising deals shaping what the data says.
Field notes and explainers from the Nibi desk - how to read a staff gauge, what a 100-year flood actually means, how snowpack becomes summer water.
For five years, spring 2019 was the mark every Lake Muskoka decision answered to. In April 2026 the lake went two centimeters higher - and the official books have not caught up yet.
Inside the freshet cycle that puts Manitoba shorelines at risk each year - and why 2026 broke the script.
The science of snow water equivalent and why it drives the season ahead.
What the numbers on the measuring stick actually tell you.
Why Central and South Texas rivers go from gravel beds to fifty-year records in an afternoon - including July 2026 on the Nueces.
One percent, every year, forever - unwrapping hydrology's most misunderstood phrase, and how it differs from our own percentiles.