Degree Celsius ℃ vs °C: Are They the Same?

℃ can look like a compact version of °C, and to a reader both normally mean Celsius. Under the hood they are different: ℃ is one Unicode character, while °C is a degree sign followed by the ordinary capital letter C.

That difference is easy to ignore until measurement text moves between spreadsheets, databases, search indexes, or older documents. Then two labels that look identical can behave differently because their code-point sequences are not the same.

℃ and °C are encoded differently

Unicode gives U+2103 a compatibility decomposition to U+00B0 followed by U+0043. NFKC follows that mapping and turns ℃ into °C; NFC does not, because the relationship is compatibility-based rather than canonical.

Why °C is the better default

For normal use, Unicode explicitly recommends the sequence °C rather than U+2103 ℃. The single-code-point form exists for compatibility with older encodings and text systems that already had a dedicated Celsius character.

That makes °C the better default for newly authored Unicode text. It is also easier to reason about: the first character is the ordinary degree sign, and the second is the ordinary Latin capital letter C.

How Celsius is written in SI notation

The SI unit name is degree Celsius, with the symbol °C. Among measurement symbols, it is a useful example of a multi-character unit symbol: when a numerical value is written with the unit, a space separates the number from the complete symbol—20 °C—but ° and C stay together.

This differs from a plane angle such as 20°. For degrees of angle, the degree sign follows the number directly. The character U+00B0 is the same; the surrounding measurement convention is different.

What normalization changes

An exact search for the two-character sequence °C does not have to match a source containing the single code point ℃. Unicode recommends treating the forms as equivalent for searching, but software only does that if its search or normalization logic has been designed accordingly.

As a result, two Celsius labels can look the same and still fail a literal string comparison. The scale is not the problem; the stored code-point sequence is.

NFC and NFKC serve different purposes. NFC preserves U+2103 because the Celsius character does not have a canonical decomposition. NFKC uses compatibility mappings, so it can convert ℃ to the sequence °C.

That can be useful for search indexes and data cleanup. It is also a real transformation: after NFKC, the original U+2103 code point is no longer present. If exact source fidelity matters, keep the original text as well as any normalized form used for matching.

The glyph will not tell you what is stored

A font can draw U+2103 almost exactly like a degree sign followed by C. Some typefaces make the one-character form slightly tighter, but that is not a dependable identification method.

If the distinction matters, inspect the code points. A visual check is useful for spotting suspicious text, not for proving which Unicode representation is present.

℃ does not name a different temperature scale

Changing ℃ to °C changes the text representation, not the physical quantity. Both are intended to represent Celsius. Converting Celsius to another scale is a separate operation that changes the numerical value.

For that measurement distinction, see Celsius °C vs Kelvin K, which explains why kelvin is written without a degree sign.

What to do with mixed Celsius data

When a dataset contains both forms, start with the purpose of the data. An archive may need to keep U+2103 exactly as it arrived. A working dataset used for search, grouping, or analysis is usually easier to maintain if the compatibility character is normalized to °C.

If both needs matter, keep the raw field and derive a normalized value for matching. That preserves the source without letting ℃ and °C split one Celsius category into two accidental buckets.

Small raised circles need separate identification

A cleanup rule should target U+2103 specifically or operate through a defined normalization policy. It should not replace every character that merely looks like a degree sign. The masculine ordinal indicator º, ring above ˚, and superscript zero ⁰ are different characters with different meanings.

If the problem is the degree mark itself rather than the Celsius compatibility character, inspect that character separately before changing the data.

Copy and paste can preserve U+2103 even when the destination font makes it look exactly like °C. If a value came from an older PDF, spreadsheet, or imported feed, checking the code points is therefore more reliable than retyping only when something looks visibly wrong.

For new text, write °C

For newly written Unicode text, use °C. Keep ℃ when you are reproducing a source faithfully or dealing with a system that specifically requires U+2103.

The key point is simple: ℃ is a compatibility representation of the same visible unit, not a second Celsius scale or a more “correct” Celsius symbol.