The Docket | Emerging Science

Lead Study

The Number Beneath the Gate

By David Kern— April 22, 2026

Feature image for article: The Number Beneath the Gate

In December 2025, engineers began designing the largest flood gate system in the world. Two miles across the mouth of Galveston Bay. Gate crest elevation: 21.5 feet. Budget: $34 billion. Construction through 2043, then fifty years of operation. Everything that follows depends on that number being right.

Four months later, a study in *Science Advances* corrected a salinity bias that had been distorting the ocean circulation models those projections rest on. The revision narrowed the uncertainty range by 79%, toward the worse end. And the revised number is still incomplete.

Lead Study

The Number Beneath the Gate

In December 2025, engineers began designing the largest flood gate system in the world. Two miles across the mouth of Galveston Bay. Gate crest elevation: 21.5 feet. Budget: $34 billion. Construction through 2043, then fifty years of operation. Everything that follows depends on that number being right.

Four months later, a study in *Science Advances* corrected a salinity bias that had been distorting the ocean circulation models those projections rest on. The revision narrowed the uncertainty range by 79%, toward the worse end. And the revised number is still incomplete.

David Kern

Fiction writer and commercial pilot known for emotionally honest stories about grief, hope, and ordinary people managing loss. His plain-spoken style has made him widely considered the best writer of his generation.

Shifting Baselines

Six findings landed this week. They cover different systems, different geographies, different journals. What they share is a specific kind of problem: each one describes a baseline that shifted while the instruments calibrated to it kept reading normal.

Fires that used to cool at night don't anymore. State temperature averages look fine while the extremes underneath them warp. An ocean methane source feeds itself through a loop no climate model tracks. Africa's forests flipped from absorbing carbon to emitting it, and the budgets built on the old assumption ran for a decade before anyone corrected the ledger.

These aren't predictions. They're measurements of where reality already diverged from the numbers we've been planning with.

Shifting Baselines

Six findings landed this week. They cover different systems, different geographies, different journals. What they share is a specific kind of problem: each one describes a baseline that shifted while the instruments calibrated to it kept reading normal.

Fires that used to cool at night don't anymore. State temperature averages look fine while the extremes underneath them warp. An ocean methane source feeds itself through a loop no climate model tracks. Africa's forests flipped from absorbing carbon to emitting it, and the budgets built on the old assumption ran for a decade before anyone corrected the ledger.

These aren't predictions. They're measurements of where reality already diverged from the numbers we've been planning with.

Wildfire Science

Wildfires Have Stopped Dying Down at Night

Fire-weather hours across North America are up 36% since the mid-1970s. California alone has gained 550 more potential burning hours. Nights are warming faster than days, and the cool, humid window that firefighters relied on to gain ground is closing. "Nights aren't what they used to be," says UC Merced fire scientist John Abatzoglou. Suppression strategy was built around a nighttime reprieve that increasingly doesn't arrive.

Ocean Feedback

A Methane Feedback the Models Don't Know

University of Rochester researchers found that bacteria in oxygen-rich surface oceans produce methane when phosphate runs low. Warming stratifies the ocean, choking off nutrient mixing from below, which starves the surface of phosphate and feeds the cycle. The feedback is real, documented, and absent from every major climate projection model currently in use.

Temperature Data

U.S. Warming Is Real but Averages Lie

Only half of U.S. states show rising average temperatures. But a PLOS Climate study drawing on 26,000-plus daily readings per state found 41 states warming in at least one part of the temperature distribution. The West is stacking extreme heat days. The North is shedding cold extremes. States in the so-called "Warming Hole" look stable on paper while the tails of their distributions quietly fatten.

Carbon Sink

Africa's Forests Now Emit More Than They Absorb

Sometime after 2010, tropical deforestation flipped Africa's forests from carbon sink to carbon source. The continent now loses roughly 106 billion kilograms of forest biomass a year. Global carbon budgets assumed these forests would keep pulling carbon down. That assumption has been wrong for over a decade, and the math built on it hasn't caught up.

Climate Target

1.5 Degrees Is Gone. Now What's the Number?

Resources for the Future harmonized 15 scenarios from eight major energy organizations for its Global Energy Outlook 2026. The conclusion is blunt: the world will blow past 1.5°C before 2050. Holding to 2°C is called "extremely challenging" and requires policy effort beyond anything currently committed. The planning anchor that organized a decade of climate policy no longer holds.

Marine Ecology

Gray Whales Are Dying in San Francisco Bay

Since 2019, gray whales have been breaking centuries-old migration patterns and turning into San Francisco Bay. Seventy carcasses have washed up nearby. Arctic warming is collapsing the benthic food web they depend on, and malnourished whales are wandering into an urban waterway where ship strikes finish what starvation started. The detour is not adaptation. It's a death spiral with witnesses.

Research Papers

Observational constraints project a ~50% AMOC weakening by the end of this century

Researchers applied ridge-regularized regression to test CMIP6 models against 19 ocean observations, then reweighted projections by observational accuracy. The corrected estimate: 51% AMOC slowdown by 2100, with uncertainty collapsed from ±37 to ±8 percentage points. South Atlantic salinity bias drove nearly half the correction.

Where was the bias hiding?

Models systematically underestimated South Atlantic salinity, a single regional error that accounted for 47% of the gap between old and corrected AMOC projections.

Is the science settled?

Potsdam oceanographer Stefan Rahmstorf called it "seriously bad news," though a rival constraint study using different observational inputs reached the opposite conclusion.

Phosphate scarcity governs methane production in the global open ocean

University of Rochester researchers resolved the "ocean methane paradox" by identifying phosphate scarcity, not oxygen absence, as the primary control on microbial methane production in sunlit surface waters. Warming-driven stratification will shrink nutrient mixing, expanding phosphate-depleted zones and feeding a methane cycle absent from current climate models.

What mechanism was missing?

Certain bacteria switch to methane production only when phosphate runs scarce, making nutrient depletion the governing variable, not dissolved oxygen levels.

Why does the model gap matter?

This ocean-to-atmosphere methane feedback is entirely unrepresented in major climate projections, meaning an amplifying cycle now has documentation but no modeling home.

Research Papers

Observational constraints project a ~50% AMOC weakening by the end of this century

Researchers applied ridge-regularized regression to test CMIP6 models against 19 ocean observations, then reweighted projections by observational accuracy. The corrected estimate: 51% AMOC slowdown by 2100, with uncertainty collapsed from ±37 to ±8 percentage points. South Atlantic salinity bias drove nearly half the correction.

Where was the bias hiding?

Models systematically underestimated South Atlantic salinity, a single regional error that accounted for 47% of the gap between old and corrected AMOC projections.

Is the science settled?

Potsdam oceanographer Stefan Rahmstorf called it "seriously bad news," though a rival constraint study using different observational inputs reached the opposite conclusion.

Research Papers

Phosphate scarcity governs methane production in the global open ocean

University of Rochester researchers resolved the "ocean methane paradox" by identifying phosphate scarcity, not oxygen absence, as the primary control on microbial methane production in sunlit surface waters. Warming-driven stratification will shrink nutrient mixing, expanding phosphate-depleted zones and feeding a methane cycle absent from current climate models.

What mechanism was missing?

Certain bacteria switch to methane production only when phosphate runs scarce, making nutrient depletion the governing variable, not dissolved oxygen levels.

Why does the model gap matter?

This ocean-to-atmosphere methane feedback is entirely unrepresented in major climate projections, meaning an amplifying cycle now has documentation but no modeling home.

River Forecast

The Colorado River's April Snowpack Forecasts Have Been Missing the Variable That Matters Most

Every April, water managers hike into the Rockies, measure snowpack, and tell 40 million people how much Colorado River water to expect. Since roughly 2000, those forecasts have consistently overpromised. The river keeps coming up short, and nobody could fully explain why.

University of Washington researchers now have an answer: spring rain, or the lack of it. Warmer, drier springs let forests and brush drink snowmelt before it ever reaches the river. That single variable accounts for 67% of the forecasting gap. The measurement everyone trusted was real. It just wasn't measuring what mattered anymore.

"Spring hasn't occurred yet," co-author Jessica Lundquist said of the April forecasting window. "Now that we know spring rain is actually more important than rain any other times of the year, we're going to have to get better at predicting what's going to happen rainwise." The pattern tracks with what hydrologists call the Millennium drought, which began around 2000 and shows no sign of ending.

River Forecast

The Colorado River's April Snowpack Forecasts Have Been Missing the Variable That Matters Most

Every April, water managers hike into the Rockies, measure snowpack, and tell 40 million people how much Colorado River water to expect. Since roughly 2000, those forecasts have consistently overpromised. The river keeps coming up short, and nobody could fully explain why.

University of Washington researchers now have an answer: spring rain, or the lack of it. Warmer, drier springs let forests and brush drink snowmelt before it ever reaches the river. That single variable accounts for 67% of the forecasting gap. The measurement everyone trusted was real. It just wasn't measuring what mattered anymore.

"Spring hasn't occurred yet," co-author Jessica Lundquist said of the April forecasting window. "Now that we know spring rain is actually more important than rain any other times of the year, we're going to have to get better at predicting what's going to happen rainwise." The pattern tracks with what hydrologists call the Millennium drought, which began around 2000 and shows no sign of ending.

Thirsty forests:

Plants get first access to snowmelt. In drier springs, vegetation consumes water that once flowed freely to city intakes and irrigation canals downstream.

Elevation gradient:

Lower basins lose the most water because earlier melt gives plants a longer growing season to intercept runoff before it reaches the river.

Current stakes:

Lake Powell sits at 24% capacity, with 2026 snowpack at record lows and a hot March accelerating exactly the mechanism this study documents.

Drought or permanent:

USGS scientists are now debating whether the Millennium drought is actually aridification, a distinction that rewrites every planning assumption on the river.

Spring drying:

Precipitation across headwater basins has dropped 14% since 2000, with lower and middle elevations absorbing the steepest declines.

Reading List

Live Science: "Nations Need to Prepare Now" — Key Atlantic Current Much Closer to Collapse Than Scientists Thought

Portmann and Rahmstorf in their own words on why the pessimistic AMOC models now look like the realistic ones.

US News / AP: Wildfires Used to Go to Sleep at Night. Climate Change Has Them Burning Overtime.

Abatzoglou and Wang walk through the data on vanishing nighttime recovery windows for fire crews across North America.

ScienceDaily: Spring Rain, Not Snowpack, Behind the Colorado River's Chronic Forecast Misses

Resources for the Future: Global Energy Outlook 2026 — Fifteen Scenarios, No Path to 1.5°C

AGU: Much of Humanity May Face Hot-Dry Extremes Five Times More Often by Century's End

TIME: Climate Change Is Making People Too Hot to Exercise — With Fatal Consequences

Reading List

Live Science: "Nations Need to Prepare Now" — Key Atlantic Current Much Closer to Collapse Than Scientists ThoughtPortmann and Rahmstorf in their own words on why the pessimistic AMOC models now look like the realistic ones.

US News / AP: Wildfires Used to Go to Sleep at Night. Climate Change Has Them Burning Overtime.Abatzoglou and Wang walk through the data on vanishing nighttime recovery windows for fire crews across North America.

Quick links

ScienceDaily: Spring Rain, Not Snowpack, Behind the Colorado River's Chronic Forecast Misses

Resources for the Future: Global Energy Outlook 2026 — Fifteen Scenarios, No Path to 1.5°C

AGU: Much of Humanity May Face Hot-Dry Extremes Five Times More Often by Century's End

TIME: Climate Change Is Making People Too Hot to Exercise — With Fatal Consequences

Past Articles

What a Decade of Climate Costs Looks Like in San Antonio

In 2023, San Antonio hit 75 days above 100°F. No president flew in. No anchors stood in anybody's yard. Insurance premiu...

The Warming Hole — What It Means When the Thermometer Says Your State Is Fine

In 2023, more than 300 Texans died from heat, the most since the state began tracking. The grid that nearly collapsed du...

The Same Warming, Two Different Dangers

Rain at 10,000 feet in the Idaho Sawtooths. Rain in the Bridger Range above Bozeman, midwinter, where snowpack is suppos...

When Two Crises Share a County but Not a Plan

Eighty-eight percent of the people who died indoors from heat in Maricopa County last year had an air conditioning unit ...

Past Articles

What a Decade of Climate Costs Looks Like in San Antonio

In 2023, San Antonio hit 75 days above 100°F. No president flew in. No anchors stood in anybody's yard. Insurance premiu...

The Warming Hole — What It Means When the Thermometer Says Your State Is Fine

In 2023, more than 300 Texans died from heat, the most since the state began tracking. The grid that nearly collapsed du...

The Same Warming, Two Different Dangers

Rain at 10,000 feet in the Idaho Sawtooths. Rain in the Bridger Range above Bozeman, midwinter, where snowpack is suppos...

When Two Crises Share a County but Not a Plan

Eighty-eight percent of the people who died indoors from heat in Maricopa County last year had an air conditioning unit ...