📄 earth science

Ice-core d-excess reveals a shift to wind-driven control of oceanic evaporation in warm climates

By reframing deuterium-excess in Antarctic ice cores as a proxy for oceanic evaporation rates, this study reveals an 800,000-year history where the control of evaporation shifts from thermodynamic temperature dependence in cold glacial periods to wind-driven dynamics in warm climates, suggesting future water cycle responses will be equally governed by atmospheric circulation.

Original authors: Ajay Ajay, Alexandre Cauquoin, Kei Yoshimura, Jeffrey Welker, Prasanta Sanyal

Published 2026-09-09

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Ajay Ajay, Alexandre Cauquoin, Kei Yoshimura, Jeffrey Welker, Prasanta Sanyal

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The story of Earth's climate is written in the air, but much of that story is hidden in the water vapor that rises from the oceans. For decades, scientists have looked to ancient ice sheets in Antarctica as a library of past climates, reading the chemical signatures trapped in layers of ice that fell as snow hundreds of thousands of years ago. One specific signature, known as deuterium excess, has long been a source of confusion. It is a subtle difference in the weight of water molecules that forms as ocean water evaporates. Traditionally, researchers used this signal to guess the temperature of the ocean where the moisture began its journey or how humid the air was at that moment. However, these two factors, along with the speed of the wind, all mix together to create the signal. Because they are so intertwined, it has been difficult to tell which one was actually driving the changes seen in the ice, leaving a gap in our understanding of how the global water cycle has responded to a warming or cooling planet.

A new study by Ajay Ajay and colleagues at the Indian Institute of Science Education and Research Kolkata, along with collaborators from the University of Tokyo and the University of Alaska Anchorage, has resolved this ambiguity by changing the question entirely. Instead of trying to separate temperature, humidity, and wind to find a single cause, the researchers realized that the deuterium excess signal actually tracks the total amount of water evaporating from the ocean surface. By treating this ancient chemical signature as a direct measure of evaporation rates, they were able to reconstruct a continuous history of how much moisture the Southern Ocean has released into the atmosphere over the last 800,000 years. This record covers multiple cycles of ice ages and warmer periods, offering the first long-term view of the very first step of the global water cycle.

The researchers began by proving that their new approach worked in the modern world. They compared the deuterium excess values calculated from modern weather data with actual measurements of ocean evaporation from satellites and ship-based instruments. They found a strong, reliable connection: where the wind was stronger and the air was drier, the deuterium excess was higher, and the rate of evaporation was faster. This relationship held true across the specific region of the Southern Indian Ocean that feeds moisture to the Antarctic ice sheet. Once this link was confirmed, they applied the same logic to the ancient ice core from the EPICA Dome C site in Antarctica. This core contains a pristine record of snowfall stretching back 800,000 years, allowing the team to translate the chemical shifts in the ice into a timeline of evaporation intensity.

What emerged from this timeline was a surprising shift in how the climate system behaves. During the coldest periods of the past, when the world was locked in deep ice ages, the amount of evaporation was tightly controlled by the temperature of the ocean surface. As the ocean warmed or cooled, the evaporation rose or fell in direct step with it. However, as the climate warmed into the transitional periods and the current warm interglacial eras, this relationship broke down. The researchers found that during these warmer times, the temperature of the ocean no longer dictated how much water evaporated. Instead, the speed of the wind became the dominant force. When the wind picked up, it swept dry air over the ocean, pulling more moisture into the atmosphere even if the ocean temperature remained steady.

This discovery reveals that the control over the Earth's water cycle is not fixed; it flips depending on the state of the climate. In cold worlds, the heat of the ocean drives the process. In warm worlds, the movement of the air drives it. The study shows that source temperature and wind speed together explain 81 percent of the changes in evaporation over the last 800,000 years, whereas temperature alone only explains 55 percent. This means that in a warming climate, the strength of the water cycle is governed just as much by atmospheric circulation and wind patterns as it is by rising temperatures. The findings suggest that future changes in global evaporation and precipitation will depend heavily on how the winds respond to a warmer planet, a factor that has been difficult to predict until now. By reading the ice not as a thermometer but as a gauge of the ocean's breath, the researchers have provided a clearer, more physical picture of how our planet's water moves.

Technical Summary: Ice-core d-excess reveals a shift to wind-driven control of oceanic evaporation in warm climates

Problem Statement
Reconstructing past oceanic evaporation rates ( $E$ ) has remained elusive because evaporation is governed by multiple, inseparable boundary conditions: sea surface temperature (SST), air-sea humidity gradients, and wind speed. While deuterium excess ( $d$ -excess) in Antarctic ice cores has long been used as a proxy for moisture-source temperature or relative humidity, its interpretation has been ambiguous. $d$ -excess responds simultaneously to SST, humidity, and wind, making single-variable reconstructions undetermined. Furthermore, traditional interpretations often focus on source conditions rather than the flux itself, limiting the ability to quantify the strength of the global water cycle's first step.

Methodology
The authors propose reframing $d$ -excess not as a tracer of source conditions, but as a direct proxy for the oceanic evaporation rate. This approach leverages the underlying physics where kinetic fractionation during evaporation is driven by the evaporative flux.

Calibration: The study establishes an empirical transfer function between $d$ -excess and evaporation rates using modern data. The authors utilized ERA5 reanalysis data to calculate $d$ -excess fields based on the Uemura et al. (2008) relationship, specifically calibrated for the Southern Indian Ocean (35°S–65°S), the primary moisture source for the EPICA Dome C (EDC) ice core.
Validation: The derived $d$ -excess field was validated against independent evaporation products (OAFlux, J-OFURO3, and HOAPS) and isotope-enabled General Circulation Models (GCMs) from the Water Isotope Model Intercomparison Project (WISOMIP). The correlation between $d$ -excess and evaporation was found to be robust ( $R^2 \approx 0.90$ ) over the global ocean and exceptionally strong ( $R^2 \approx 0.95$ ) over the EDC source region.
Reconstruction: This transfer function ( $E = 0.266d^2 + 2.155d + 17.40$ ) was applied to the continuous 800,000-year EDC ice-core record to reconstruct past evaporation anomalies ( $\Delta E\%$ ).
Driver Attribution: To disentangle the controls on evaporation, the study analyzed the relationship between reconstructed evaporation, source temperature ( $\Delta T_{source}$ ), and wind speed. Wind speed was proxied by the EDC dust-flux record (Lambert et al., 2012), which serves as an external, non-circular indicator of atmospheric circulation and ventilation. The authors computed residuals of evaporation independent of temperature to isolate the wind-driven signal.

Key Results

Continuous Evaporation Record: The study produces the first continuous 800,000-year history of oceanic evaporation at the Antarctic moisture source.
Regime-Dependent Control: The control mechanism governing evaporation shifts fundamentally with climate state:
- Cold Glacial Regimes ( $\Delta T_{source} \le -3.5^\circ$ C): Evaporation is thermodynamically controlled. Source temperature alone explains 57% of the variance ( $R^2 = 0.57$ ). In these conditions, the boundary layer is near-saturated, making the humidity gradient insensitive to wind; thus, evaporation tracks temperature changes.
- Warm and Transitional Regimes ( $\Delta T_{source} \ge -1.0^\circ$ C and $-3.5^\circ$ C $< \Delta T_{source} < -1.0^\circ$ C): Evaporation becomes dynamically controlled. Source temperature explains only 10–23% of the variance. Instead, wind-driven transport (proxied by dust flux) becomes the dominant factor. In warm climates, the atmosphere holds more moisture, leading to a saturated boundary layer where evaporation is limited by the rate at which wind advects vapor away (ventilation).
Decoupling Events: During interglacial-to-glacial transitions, peak source temperature and peak evaporation are decoupled. For example, during the last interglacial, source temperature peaked at 122.8 ka, while reconstructed evaporation continued to rise, reaching a maximum at 104.2 ka (an 18.7 kyr lag). This lag is driven by increasing wind speeds (dust flux) even as the source cools.
Variance Explanation: Across the full 800-kyr record, the combination of source temperature and wind (dust flux) explains 81% of the evaporation variance, compared to 55% for temperature alone.

Significance and Claims
The paper claims to resolve the long-standing ambiguity of $d$ -excess by demonstrating its utility as a quantitative measure of oceanic evaporation rather than just a source-condition tracer. The primary significance lies in identifying a "thermodynamic-to-dynamic" transition in the hydrological cycle:

In cold climates, evaporation is driven by temperature (thermodynamics).
In warm climates, evaporation is driven by atmospheric circulation and wind (dynamics).

The authors conclude that atmospheric circulation is a first-order control on the strength of the hydrological cycle, particularly in warm states relevant to future climate warming. This implies that the response of the water cycle to future warming cannot be inferred from sea surface temperature or humidity gradients alone; atmospheric circulation patterns will play a governing role. The reconstruction, now cast in physical units, allows for direct testing against ocean-atmosphere water cycle theories and models.

More like this