How this instrument works
Kelvin is an SI base unit; Celsius is a scale built directly on top of it. Since 20 May 2019 one kelvin has been fixed by assigning Boltzmann's constant an exact value of 1.380649 × 10⁻²³ joules per kelvin, tying temperature to energy rather than to any particular substance. Celsius then follows from one defining equation printed in the SI Brochure: t/°C = T/K − 273.15. Subtraction is a definition here, so this conversion sheds no accuracy at all.
Where does 273.15 come from? From 1954 until 2019, kelvin was anchored to water's triple point, assigned exactly 273.16 K — chosen so each degree kept its familiar size, with 100 steps between melting ice and boiling water at standard pressure. Ice melts 0.01 K below that triple point, which places 0 °C at 273.15 K and fixes an offset that survived redefinition untouched. Anders Celsius published his hundred-step scale in 1742 running backwards, with 0 for boiling and 100 for freezing; colleagues flipped it within years of his death. William Thomson, later Lord Kelvin, argued in 1848 for an absolute scale whose zero was thermodynamic rather than conventional.
Each unit keeps its own working territory. Physicists, cryogenic engineers and astronomers quote kelvin because ratios matter to them: 77.36 K for liquid nitrogen, 2.725 K for cosmic microwave background, 5772 K at our Sun's surface. Lighting and photography borrow kelvin too, though 2700 K printed on lamp packaging describes blackbody colour rather than warmth anyone could feel. Celsius runs everything human-scaled almost everywhere outside America — forecasts, ovens, fevers, cold-chain logs, concrete cure schedules. One datasheet may quote storage limits in Celsius and semiconductor junction limits in kelvin on facing pages.
- Type your absolute temperature into the Kelvin (K) field; 300 is preloaded as a starting point.
- Read the Celsius (°C) line beneath it — it recomputes with every keystroke, no button to press.
- Working backwards from Celsius? Add 273.15 to your figure, enter that into Kelvin (K), and the Celsius (°C) line should hand back what you began with.
- Cryogenic figures need resolution, so the Kelvin (K) field steps in thousandths; paste finer values straight in if your instrument reports them.
- Nothing below 0 K exists, so a negative entry signals a transcription slip rather than a very cold sample.
Worked example — what 300 K actually feels like
Physics problems say "take room temperature as 300 K" so often that it starts to sound like a definition. Enter 300 into Kelvin (K) and this sheet returns 26.85 °C, which is no comfortable office at all — that is a warm summer afternoon, several degrees above where most thermostats settle.
For contrast, 20 °C is 293.15 K and 25 °C is 298.15 K. Shortcut 273.15 to a flat 273 and that same input reads 27.00 °C, high by 0.15. Such a gap vanishes in weather talk yet shows up plainly in thermistor calibration, gas-law volumes, and any reaction-rate figure.
Questions
Is the 273.15 offset exact or rounded?
Exact. SI defines Celsius temperature by t/°C = T/K − 273.15, so that figure belongs to a definition and carries no uncertainty of its own. Convert either direction and nothing is lost — any imprecision in your answer arrived with your input reading or came from how many decimals you chose to display.
Can I just subtract 273 and move on?
Only if 0.15 °C sits beneath your tolerance. Meteorologists and cooks lose nothing by it. Anyone calibrating a platinum resistance thermometer, computing gas-law volumes, or applying Arrhenius kinetics should keep all five digits: 0.15 K against a 300 K baseline is 500 parts per million, larger than many instruments' own stated uncertainty.
Do I subtract 273.15 from a temperature difference too?
No, and this trips people up more than anything else about these two scales. Celsius degrees and kelvins are identical in size, so an interval of 12 K is an interval of 12 °C with no shift applied. Offsets belong to points on a scale. Differences, ramp rates such as °C per minute, and coefficients like W/(m·K) or J/(kg·K) carry straight across untouched.
Should I write °K or say degrees Kelvin?
Neither. In 1967 the 13th CGPM dropped that degree sign, making "kelvin" that unit's full name and K its symbol — you write 300 K and say "three hundred kelvin", lowercase in prose, capital K in symbol form. Celsius keeps both its degree sign and its capital C because it honours a person: 26.85 °C, with a space before that symbol.
What is absolute zero in Celsius?
Precisely −273.15 °C, which is 0 K by construction. Thermodynamic temperature has a floor, and Celsius merely relabels it. Laboratories have crept astonishingly close: nuclear spin systems in rhodium were driven to roughly 100 picokelvin, about one ten-billionth of a degree above that floor, though bulk matter around them stayed far warmer.
Is the 2700 K on a light bulb box a real temperature?
That is colour, not heat. Correlated colour temperature names whichever blackbody a lamp's light most resembles: 2700 K for warm incandescent tint, 4000 K neutral white, 6500 K overcast daylight. Running that figure through this converter yields 2426.85 °C, which says nothing useful about an LED whose housing sits near 60 °C. Confusingly for shoppers, lower numbers look warmer.