📄 earth science

Asymmetric day–night warming across Pakistan, 1980–2025: diurnal temperature range compression in the irrigated Indus plains and expansion in the Karakoram

This study reveals that Pakistan's warming from 1980 to 2025 is highly asymmetric, characterized by a compression of the diurnal temperature range in the irrigated Indus plains due to rapidly rising minimum temperatures, while the Karakoram region experiences an expansion of the range as maximum temperatures increase faster than minimums.

Original authors: Ibrar Junaid, Shahood Sajid

Published 2026-10-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ibrar Junaid, Shahood Sajid

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

Heat does not arrive at the same time of day everywhere, nor does it leave at the same speed. For decades, scientists have tracked how the Earth's surface warms, noticing a distinct pattern in how the temperature swings between the bright heat of noon and the cool quiet of night. This daily swing, known as the diurnal temperature range, is not just a number on a chart; it is a vital rhythm for the living world. When nights stay too warm, crops struggle to rest and recover, and people find it harder to cool down after a scorching day. In many parts of the world, nights have been warming faster than days, shrinking the gap between the two. However, this pattern is not universal. In some places, the sun is getting stronger during the day while the nights remain relatively stable, or the two extremes are shifting in completely different directions. Understanding these local rhythms is crucial because a single average temperature for a whole country can hide the specific dangers that farmers and doctors face.

A new study looking at Pakistan over the last forty-six years reveals that the country is not warming in a single, uniform way. Instead, it is split into two distinct zones that are experiencing opposite changes in their daily temperature cycles. The researchers, using detailed computer models of the atmosphere that blend satellite data with ground observations, analyzed the daily high and low temperatures across the nation from 1980 to 2025. They found that in the vast, irrigated river plains where most of the country's food is grown, the nights are getting significantly warmer while the days are barely changing. This has caused the daily temperature gap to shrink dramatically. In contrast, in the high, rugged mountains of the north, the days are heating up much faster than the nights, causing the daily temperature gap to widen.

The irrigated plains, covering the provinces of Punjab and Sindh, tell a story of a quiet but profound shift. Here, the minimum temperature—the lowest point reached just before dawn—has risen by about three-tenths of a degree Celsius every ten years. Meanwhile, the maximum temperature, the peak heat of the afternoon, has remained almost exactly the same. Because the nights are getting warmer but the days are not, the difference between the two has compressed by nearly three-tenths of a degree per decade. Over the course of the study, this adds up to a loss of more than one degree of daily temperature variation. This compression is most intense during the monsoon and the months that follow, when the region is wet and humid. It is a phenomenon that happens almost exclusively at night. The study suggests this is driven by a combination of factors: the massive irrigation systems that keep the soil moist, which cools the air during the day but traps heat at night, and a thickening of the atmosphere with water vapor and clouds that blocks the sun's heat from reaching the ground while trapping the warmth below.

In the high mountains of Gilgit-Baltistan, the story is reversed. Here, the maximum daytime temperature is rising much faster than the nighttime low. The days are getting hotter, while the nights are warming at a slower pace. This has caused the daily temperature range to expand, or grow wider, by about one-tenth of a degree every ten years. The researchers link this to a loss of low-hanging clouds in the region. With fewer clouds to block the sun, the ground bakes more intensely during the day, but without a corresponding increase in the heat-trapping moisture that keeps nights warm, the nights do not keep up. This pattern is particularly important for the region's glaciers and snowpack, because ice melts primarily when the sun is out. A warming trend driven by hotter days means more melt energy is delivered to the ice than if the warming were driven by warmer nights.

The study also looked at how these changes affect the frequency of extreme weather. Across the whole country, the number of "warm nights"—nights that stay uncomfortably hot—is increasing much faster than the number of "warm days." In the plains, the percentage of nights that feel like a tropical summer has more than doubled since the early 1980s, rising from less than eight percent of all nights to more than twenty percent. In the mountains, however, the trend flips, with hot days becoming more common than hot nights. This distinction matters deeply for agriculture and health. In the plains, where wheat and rice are grown, crops need cool nights to finish filling their grains. The rapid rise in nighttime temperatures shortens this critical growing period, threatening harvests. For people, warm nights prevent the body from recovering from the heat of the day, increasing the risk of heat-related illness and death.

The researchers were careful to note that their findings rely on a mix of satellite data and ground measurements, and that in the remote, high-altitude regions, the data is less direct than in the populated plains. They also found that the data from the mountains is not yet confirmed by a dense network of local weather stations, meaning the mountain results are a strong signal from the models that needs future verification. However, the pattern in the plains is robust and confirmed by independent station records. The study concludes that looking at the average temperature for Pakistan as a whole is misleading. It hides the fact that the country is experiencing two different kinds of warming: one that chokes the crops and traps heat in the valleys, and another that bakes the high peaks. To protect the country's food supply and its people, planners and scientists must stop relying on a single average number and start paying attention to the specific rhythm of the day and the night.

Technical Summary: Asymmetric Day–Night Warming Across Pakistan (1980–2025)

Problem Statement
Global warming is rarely symmetric between day and night, leading to regionally specific changes in the Diurnal Temperature Range (DTR = $T_{max} - T_{min}$ ). While early syntheses suggested nights warmed faster than days (DTR compression), later updates indicated a global reversal where daytime warming accelerated. However, existing studies on Pakistan rely on sparse station networks that end in 2017 or earlier, often fail to distinguish between opposing regional regimes, and lack robust statistical controls for serial correlation and multiple testing. Furthermore, the physical drivers of DTR changes—such as irrigation, aerosol dimming, and cloud cover—remain poorly quantified for the country's diverse agro-ecological zones. This study addresses these gaps by analyzing 46 years of data to determine if Pakistan exhibits a single national warming signature or distinct, opposing regimes.

Methodology
The study utilizes 46 years (1980–2025) of daily maximum ( $T_{max}$ ) and minimum ( $T_{min}$ ) temperature data from the ERA5 reanalysis (0.25° grid) covering Pakistan and its six first-order administrative regions.

Statistical Framework: Trends were assessed using the Mann–Kendall test with the Hamed–Rao serial-correlation correction to account for autocorrelation. Trend magnitudes were estimated via Sen's slope. To handle the multiple comparisons inherent in analyzing seven regions across eight seasons and three variables, the Benjamini–Hochberg procedure was applied to control the false discovery rate (FDR) at 5%.
Data Validation: Results were cross-validated against the CRU TS 4.10 gridded dataset and the HadEX3 station-based extremes dataset. A specific homogeneity check was performed to identify and discard data artifacts, notably a discontinuity in an alternative archive (Open-Meteo) caused by a change in the underlying blended product in 2017.
Physical Drivers: Detrended Pearson correlations were used to link DTR variability to auxiliary ERA5 fields: cloud cover, total column water vapour (TCWV), and surface solar/thermal radiation.
Elevation Analysis: Grid cells were stratified into five elevation bands to assess elevation-dependent warming (EDW) and changes in the near-surface environmental lapse rate.

Key Results
The analysis reveals that Pakistan contains two opposing diurnal warming regimes, which are masked when averaging national data:

Irrigated Indus Plains (Punjab and Sindh):
- DTR Compression: $T_{min}$ rose significantly (+0.33 and +0.28 °C decade⁻¹, respectively), while $T_{max}$ showed no significant change (+0.04 and +0.02 °C decade⁻¹).
- Magnitude: DTR compressed by approximately 0.28 °C decade⁻¹, totaling roughly 1.2 °C over the 46-year record.
- Seasonality: Compression is strongest during the monsoon (JJA) and post-monsoon (SON) seasons and is absent in winter/early spring.
- Drivers: This regime is associated with declining surface solar radiation (dimming), rising water vapour, and increased downward thermal radiation. The suppression of $T_{max}$ aligns with the expansion of irrigation, though the reanalysis captures this via assimilated observations rather than direct model parameterization.
Karakoram–Himalaya (Gilgit-Baltistan):
- DTR Expansion: $T_{max}$ rose faster (+0.48 °C decade⁻¹) than $T_{min}$ (+0.34 °C decade⁻¹), resulting in a DTR widening of +0.11 °C decade⁻¹.
- Drivers: This expansion correlates with a loss of low cloud cover and increased surface solar radiation, particularly in the Rabi season, without a compensating increase in nocturnal thermal trapping.
- Validation: This finding relies solely on ERA5, as station-based products (HadEX3) lack usable coverage in the high mountains and show uniform negative trends due to interpolation from distant plains stations.
National and Regional Averages:
- The national annual DTR trend is weak (−0.09 °C decade⁻¹) and statistically insignificant after FDR correction, effectively concealing the opposing regional signals.
- Warm Extremes: Nationally, warm nights (TN90p) increased from 7.8% to 20.5% of nights, outpacing warm days (TX90p), which rose from 8.1% to 15.4%. In the mountains, this ordering reverses.
- Elevation Dependence: Warming is elevation-dependent but asymmetric: $T_{max}$ warming increases monotonically with elevation (from +0.08 to +0.51 °C decade⁻¹), while $T_{min}$ warming is flat with height. Consequently, the near-surface environmental lapse rate flattened by 0.062 °C km⁻¹ decade⁻¹.

Key Contributions

Resolution of Spatial Heterogeneity: The study demonstrates that a single national DTR metric is misleading for Pakistan, as it averages out a significant compression in the agricultural heartland and an expansion in the high mountains.
Methodological Rigor: The application of serial-correlation corrections and false-discovery-rate control provides a more robust statistical basis for trend detection than previous national studies.
Data Quality Control: The paper identifies and discards a spurious "nocturnal warming" signal in mountain regions caused by a 2017 discontinuity in a specific data archive, emphasizing the need for homogeneity testing in complex terrain.
Independent Verification: The DTR compression in the plains is independently reproduced by the HadEX3 station-based dataset, confirming the reanalysis result for data-rich areas.

Significance and Claims
The authors claim that relying on mean temperature alone misrepresents the warming reality in Pakistan, with distinct consequences for different sectors:

Agriculture: In the irrigated plains, the warming is almost entirely nocturnal. Since warm nights shorten grain-filling periods and increase respiratory losses in wheat, the lack of daytime warming does not mitigate the yield risks associated with rising $T_{min}$ .
Health: The rapid increase in warm nights (TN90p) poses a significant risk to human health by limiting physiological recovery from daytime heat, a factor often overlooked in heat-health planning focused on $T_{max}$ .
Cryosphere: In the Karakoram, the expansion of DTR implies that warming is concentrated in daytime hours. Since melt at high elevations occurs primarily during daylight, this daytime-driven warming may accelerate glacier melt more than a mean-temperature metric would suggest.

The paper concludes that impact assessments, agricultural advisories, and heat-health early warning systems in Pakistan must be built on separate analyses of $T_{min}$ and $T_{max}$ rather than mean temperature.

Technical Summary: Asymmetric Day–Night Warming Across Pakistan (1980–2025)

More like this