How Heat Shaped My Running Data: The Chennai Temperature Effect

Chennai summers are unforgiving. Over 10 weeks of easy running (April–June 2026), I discovered that heat wasn't just uncomfortable—it was a primary driver of my cardiovascular response, efficiency trends, and recovery patterns. This analysis reveals how 30°C+ temperatures acted as a consistent confounder that rewrote what "normal" performance looked like.

The Heat Profile

Out of 17 easy runs analyzed, 15 were conducted at 30°C or hotter. The temperature landscape:

Average Temperature
30–31°C
Range
29–34°C

Heat Exposure Across Runs

30 C+ runs: 15 of 17 (88.2%) Below 30 C: 2 of 17 (11.8%)

This wasn't seasonal variation—it was the baseline. Every run lived in the heat envelope, meaning I couldn't simply compare one run to another without accounting for temperature as a primary factor.

Heat's Impact on Heart Rate and Aerobic Decoupling

The most direct heat signature appeared in aerobic decoupling—the divergence between power output and heart rate within a run. In cooler, controlled environments, aerobic decoupling typically stays below 5%. My data told a different story.

Weighted Avg Temperature vs Average HR (Per Run)

Weighted average temperature versus average heart rate by run Each dot is one run with available temperature. X axis is weighted average temperature in Celsius from 29 to 32. Y axis is average heart rate from 143 to 155 beats per minute.
Runs plotted: 0 Each dot = one run Dashed line = linear trend

Now plotting all runs that have temperature data, using the notebook's exact weighted-temperature formula from each run's temperature distribution. Hover points to see each run's date, average temperature, and average HR.

5 out of 18 runs crossed the 8% decoupling threshold, a marker of significant cardiovascular drift. Heat was present in all of them.

Decoupling Severity by Key Runs

2026-05-02 (32 C)
9.8%
2026-05-07 (30 C)
8.4%
Healthy target line
5.0%

Bars are scaled to a 10% chart max to show how far each run drifted from typical easy-run stability.

High-Decoupling Heat Runs

What's telling: these runs weren't harder or longer than others. They were hotter. The body's thermoregulatory burden increased baseline HR by 5–15 bpm, and that cost compounded over time.

Power-Per-Heartbeat as the Adaptation Signal

Despite heat's interference, one metric cut through the noise: steady power-per-heartbeat improved +8.1% across the 10-week timeline. This is the cleanest aerobic efficiency signal available in the dataset.

Why this matters: Power-per-heartbeat corrects for the absolute HR elevation caused by heat. It answers: for every beat, am I generating more mechanical work? The +8.1% gain suggests I was actually becoming more efficient—heat made the run harder, but I adapted to it.

Signal vs Noise Snapshot

Power per heartbeat
+8.1%
Steady easy pace
+1.1%

Efficiency improved clearly, while pace stayed nearly flat under similar heat load.

Meanwhile, steady-state easy pace changed only +1.1% (essentially flat), and steady HR rose as expected in warm conditions. These metrics masked the true adaptation happening underneath: my aerobic engine was getting stronger, even as the environment pushed against it.

The Heat-Fatigue Interaction

Heat didn't just elevate HR; it accelerated fatigue perception. Fatigue resilience scores (a composite of HR drift and recovery metrics) clustered in the 52–79 range, with the lowest scores appearing in runs at 32°C+.

Fatigue Resilience (Cool ~29°C)
~70–75
Fatigue Resilience (Hot 32°C+)
~55–65

Heat and Fatigue Timeline

2026-05-1431 C
Decoupling: 6.5%Resilience: 59/100
2026-05-2832 C
Zone 4 time: 41.8%Heat strain: High
2026-06-3030 C
Avg HR: 143 bpmResilience: 62.5/100

Longer runs in hotter conditions repeatedly showed stronger HR strain and lower resilience.

The interpretation: heat consumed additional recovery capacity. A 90-minute run felt more fatiguing at 32°C than a similar effort at 29°C, even though the mechanical work was identical.

Methodology: Controlling for Heat in Future Analysis

The key takeaway for my training: I should stratify all comparisons by temperature bands. A run at 29°C is fundamentally different from one at 32°C, and treating them as equivalent introduces noise.

Going Forward

The Bigger Picture

Heat is often dismissed as "just discomfort." In this dataset, it was a primary independent variable—one that elevated HR by 10–15 bpm, doubled fatigue resilience recovery time, and dominated aerobic decoupling trends. Yet beneath it, the +8.1% power-per-heartbeat gain suggests real adaptation.

Chennai's heat isn't a bug in my training data. It's the operating environment. The insight isn't to avoid it, but to quantify it, control for it, and use it to calibrate what "good" performance actually looks like when you're running in 30°C heat every week.