Scientific notation is a way of writing numbers as a coefficient multiplied by an integer power of ten. In its normalized form, a nonzero real number is written as
where is the coefficient, also called the significand, and is an integer exponent. For example, and . The notation is widely used in science, engineering, and computing because it separates a number’s significant digits from its scale.
Structure and normalization
Scientific notation builds on the decimal number system: multiplying by ten shifts each digit one place toward a higher place value, while dividing by ten shifts it one place toward a lower place value.
The two components have distinct roles:
- The coefficient gives the digits and the sign of the number.
- The exponent gives the power of ten by which the coefficient is multiplied.
In normalized decimal scientific notation, exactly one nonzero digit appears before the decimal point. For negative numbers, the normalization condition applies to the coefficient’s absolute value:
A number can have many equivalent coefficient–exponent representations:
Only is normalized under the usual convention. This restriction gives every nonzero real number a unique coefficient and exponent as numerical values, although its written form may include additional trailing zeros to communicate precision.
Zero cannot satisfy . It is normally written simply as , though a format requiring an exponent may display it as, for example, .
The name standard form is used for scientific notation in some educational systems, but elsewhere that phrase denotes other mathematical conventions.
Converting between forms
To convert an ordinary decimal numeral into normalized scientific notation:
- Place the decimal point immediately after the first nonzero digit.
- Count how many places the point has moved.
- Use a positive exponent if the point moved left, and a negative exponent if it moved right.
For example,
because the decimal point moves seven places left, while
because it moves six places right.
To convert back, multiply the coefficient by the indicated power of ten:
A negative exponent does not make the number negative. It indicates division by a positive power of ten:
Thus is positive, whereas is negative.
Relationship to logarithms
For a nonzero number , the normalized exponent can be obtained using a base-ten logarithm:
where the floor operation means taking the greatest integer not exceeding the argument.
Consequently,
The exponent therefore identifies the interval between consecutive powers of ten containing the number’s magnitude.
Arithmetic
Scientific notation makes multiplication and division particularly convenient because powers of ten can be handled separately from coefficients.
Multiplication and division
To multiply, multiply the coefficients and add the exponents:
For example,
The last step restores normalized form.
To divide, divide the coefficients and subtract the exponents:
For example,
Addition and subtraction
Numbers must first be expressed with the same power of ten:
Adding or subtracting the exponents does not perform addition or subtraction of the original numbers.
Subtraction can produce a result with a substantially smaller exponent:
In finite-precision computing, subtracting nearly equal approximate values can expose errors already present in their digits, an issue studied in numerical analysis.
Powers
For an integer ,
with a nonzero base required when is negative. For example,
As with multiplication and division, the coefficient may need renormalization afterward.
Significant figures and uncertainty
Scientific notation is useful for displaying significant figures, especially when trailing zeros would otherwise be ambiguous. For example,
have the same mathematical value. In the usual convention for reporting measurements, however, they indicate one, two, and three significant figures respectively.
The ordinary numeral does not by itself clearly indicate which trailing zeros are significant. Scientific notation makes that distinction explicit.
Scientific notation does not establish that a value is approximate. An exactly defined quantity may also be written this way. In the International System of Units, the speed of light in vacuum is exactly
Nor does a count of significant figures completely describe measurement uncertainty. An explicit uncertainty provides information that the final displayed digit alone cannot supply. A value may, for example, be reported as
with the meaning of the uncertainty specified separately.
Rounding can change the exponent. Rounding to three significant figures gives
not a normalized coefficient beginning with .
Related notations
Engineering notation
Engineering notation restricts the exponent to a multiple of three. For nonzero values, the coefficient conventionally satisfies
For example,
This aligns naturally with SI prefixes such as kilo, milli, micro, and nano. Scientific notation instead keeps the coefficient’s absolute value below ten.
E notation
Calculators, programming languages, and data files often replace “” with an E or e followed by the exponent:
6.02E23
1.5e-6
These represent and , respectively. The letter is an exponent marker in this syntax, not Euler’s number. Accepted syntax varies among software systems.
Floating-point representation
Scientific notation is closely related to floating-point representation, in which a number is stored using a sign, a significand, and an exponent. Many computers use base two internally rather than base ten, so their stored representation resembles binary scientific notation.
A displayed decimal value need not be represented exactly in a binary format. For example, is exactly one tenth as a mathematical expression, but common finite binary floating-point formats can store only an approximation to it. Writing a value in scientific notation does not itself increase computational precision or eliminate rounding error.
Scale, comparison, and limitations
For normalized positive numbers, exponents can be compared first: the number with the larger exponent is larger. If their exponents agree, their coefficients determine the ordering. Negative numbers require attention to sign, since a greater magnitude then means a smaller numerical value.
The exponent also helps describe orders of magnitude, though conventions differ. Some uses classify numbers by the lower bounding power of ten; others use the nearest power of ten. A statement about order of magnitude therefore need not be identical to a statement about the normalized exponent.
Scientific notation is particularly useful for quantities spanning wide ranges, such as atomic dimensions, astronomical distances, and very large counts. It also makes multiplicative comparisons straightforward: is times .
The notation remains a representation, not an approximation method. A finite decimal coefficient multiplied by a power of ten always represents a rational number. An irrational number cannot be written exactly with a finite decimal coefficient, though it can be approximated in scientific notation or represented using an exact symbolic coefficient. Scientific notation also does not replace units: metres and seconds share a numerical scale but describe different physical quantities.