Lesson 1.3Lesson 1.3 · The Tools of Climate Analysis
Degree-Days and Climate Data
A peak temperature is one bad afternoon; a degree-day is the whole year of heat, counted.
Same worst day, different years
Chennai and Bengaluru both hit roughly 33 °C on a bad afternoon. Yet Chennai sits above comfort nearly all year while Bengaluru only drifts over for a few spring weeks. Same worst day, completely different years of work for the cooling system. A peak temperature is blind to this — it sees one hour and nothing else. The degree-day sees the whole year: it measures how far above comfort the air sits and for how long, and adds it all into one honest number.
A CDD figure with no base temperature attached is a number with no units. Always write the base.
Heat times time, as one number
A degree-day collapses two things — how much too warm and how long — into a single figure. First pick a base temperature: the threshold above which you reckon cooling is needed. Then, for each day the average sits above it, count the degrees of exceedance and sum them across the year.
CDD (cooling degree-days) — degrees × days *above* the base → the annual cooling burden. HDD (heating degree-days) — degrees × days *below* the base → the annual heating burden.
The trick is that 120 degree-days can be 2 °C over for 60 days, or 8 °C over for 15 days — "how much" and "how long" become interchangeable and comparable. Most of India is firmly CDD-dominated; only the cold zone and the north-Indian winter generate any real HDD. One caution that the West gets wrong for us: the base temperature is a *choice*. India commonly uses about 26 °C for cooling — warmer than the Western ~18 °C — because of the wider adaptive comfort band you met in Lesson 0.2. Always state your base; a CDD figure without its base is meaningless.
The accumulator — mild peaks, monstrous totals
Picture the interactive accumulator that drives this lesson: pick a city and a cooling base, and each month's CDD stacks as a bar while the annual total climbs; HDD shows below the line for the cold months. The lesson it teaches at a glance is counter-intuitive — a city with a *mild* peak can rack up an enormous annual total simply by never cooling down.
Chennai's monthly bars never touch zero: it stays above 26 °C even in its "cool" months, so it accumulates roughly 1,200 CDD a year. Bengaluru's bars vanish for half the year, leaving only about 60 CDD. That is a near-twentyfold difference — from two cities with almost identical peak temperatures. The peak sizes the worst hour; the total sizes the year. (The temperatures in the widget are teaching values chosen to make the point clearly; real project work uses IMD normals for the specific station.)
The peak is the worst afternoon. The degree-day is the whole year. Bill the year, not the afternoon.
Where to get honest Indian climate data
A degree-day is only as good as the temperatures fed into it, so know your sources:
IMD — the India Meteorological Department's official temperature, humidity and rainfall normals (the 1991–2020 period), across hundreds of stations. The starting point for any Indian site. ISHRAE Weather Data — design weather files for around 60 cities, built for energy simulation; the professional standard in India. NBC / SP 41 — climate-zone definitions and design temperatures, for code compliance. TMY / EPW files — hourly typical-year data, the input for serious building-energy simulation.
Treat a single memorable year and phone weather-apps as *indicative only*. Always note the data period and the station: an airport-edge station can read several degrees cooler than the dense city centre it supposedly represents — the urban heat island — and that gap quietly distorts your numbers if you ignore it.
Three altitudes on the same idea
Read the band that fits you — or all three.
Your electricity bill — not the weather forecast — is the honest measure of your cooling burden. A city that is mildly warm all year round can quietly cost *more* to cool than one that has a few brutal weeks and then settles down, because the mild city never lets the building rest. If your AC seems to run forever even though it "never gets that hot here," that is the degree-day total talking, not the peak.
Use CDD to justify envelope spend and to benchmark performance as kWh per CDD. A high-CDD site rewards shading and insulation *every single day*, so the payback is fast and generous; a low-CDD site has a shorter cooling season and a longer payback, so spend proportionately rather than over-specifying. Put the CDD figure and its base temperature in the brief alongside the comfort target, so the envelope budget is anchored to a number that survives value-engineering.
The definition is a clipped sum: CDD = Σ max(T_i − T_b, 0) over the year, with the daily mean T_i and base T_b (default T_b = 26 °C for India). A quick monthly estimate is CDD_month ≈ max(T_m − T_b, 0) · N, with T_m the month mean and N its days. Chennai month mean 30.5 °C, base 26, 31 days → (30.5 − 26) × 31 = 4.5 × 31 = 139.5 degree-days. A Bengaluru month at 24 °C → max(24 − 26, 0) = 0, contributing nothing. Note the monthly-mean method *undercounts* versus a daily sum (a month averaging 25 °C still has hot days above the base that the mean hides), so use daily data where you can. HDD just flips the sign: HDD = Σ max(T_b − T_i, 0).
“The hottest city needs the most cooling, so design for the highest peak.”
Run the method yourself
Turn your own city into a number before the next lesson.
- 1Pull the 12 monthly mean temperatures for your city from the IMD normals (1991–2020).
- 2Choose and write down a cooling base — start with 26 °C.
- 3Compute
max(T_m − 26, 0) · Nfor each month and sum them to an annual CDD. - 4Compare your total against a contrasting Indian city (a coast vs a hill station), then link the number to one concrete design decision — say, whether shading earns its cost here.
↳ Use the worksheet below to record your answers.
Take it with you
A peak is a snapshot; a degree-day is the film
We can now locate the sun (1.1), read the moisture (1.2) and total the heat (1.3). One thing is still missing: a single map that says which passive strategies will actually work on this site and which won't. The next lesson, the Givoni bioclimatic chart, overlays a city's temperature-humidity data onto zones of viable strategy — the module's everything-in-one-decision diagram.