Summary
The Gauss circle problem asks how many integer lattice points lie inside or on a circle of radius $r$ centered at the origin. If this number is denoted by $N®$, then equivalently
$N®={(m,n)\in\mathbb Z^2:m^2+n^2\le r^2}$.
Since a circle has area $\pi r^2$ and, on average, each unit square of the integer grid contains one lattice point, it is natural that
$N®\approx \pi r^2$.
The central question is therefore how large the error term
$E®=N®-\pi r^2$
can be.
Historical progress
Gauss himself established an error of essentially linear size, $E®=O®$. Later, Hardy and Landau proved that the error cannot be as small as $O(r^{1/2})$; in particular,
$|E®|\neq o\left(r^{1/2}(\log r)^{1/4}\right)$.
The famous conjecture is that the true order is almost square-root size:
$|E®|=O_\varepsilon(r^{1/2+\varepsilon})$
for every $\varepsilon>0$.
The best bound stated in the article is
$|E®|=O(r^{131/208})$
where
$\frac{131}{208}\approx0.6298$.
Thus there remains a significant gap between the known upper exponent $0.6298\ldots$ and the conjectured exponent $1/2$.
Connection with sums of two squares
The problem has a beautiful connection with number theory. Let $r_2(n)$ denote the number of representations of $n$ as a sum of two squares, counting signs and order. Then
$N®=\sum_{n=0}^{r^2}r_2(n)$.
So the geometric problem of counting lattice points becomes a problem about the arithmetic structure of integers and their representations as
$n=m^2+k^2$.
There is also an exact formula involving floor functions, obtained using Jacobi's two-square theorem.
Why the problem is difficult
The main difficulty is that the boundary of a circle does not align with the integer grid. As $r$ changes continuously, $N®$ changes only when the circle crosses a lattice point, whereas $\pi r^2$ changes smoothly. Consequently, the error $E®$ oscillates considerably.
For example, at $r=5$ there are $81$ lattice points, while
$\pi(5)^2=25\pi\approx78.54$,
giving an error of about $2.46$. Small changes in the radius can produce noticeably different errors.
Bigger picture
The Gauss circle problem is a fundamental example of a lattice-point counting problem in analytic number theory. It connects geometry, Fourier analysis, exponential sums, Bessel functions, and the arithmetic of sums of two squares. It also has higher-dimensional analogues involving lattice points inside spheres and other shapes.
In one sentence: The Gauss circle problem asks how closely the number of integer points inside a large circle follows its area $\pi r^2$; although we know the error is substantially smaller than $r$, proving the conjectured near-$r^{1/2}$ error remains an important open problem.
Read the full Wikipedia article
The Gauss circle problem asks how many integer lattice points lie inside or on a circle of radius $r$ centered at the origin. If this number is denoted by $N®$, then equivalently
$N®={(m,n)\in\mathbb Z^2:m^2+n^2\le r^2}$.
Since a circle has area $\pi r^2$ and, on average, each unit square of the integer grid contains one lattice point, it is natural that
$N®\approx \pi r^2$.
The central question is therefore how large the error term
$E®=N®-\pi r^2$
can be.
Historical progress
Gauss himself established an error of essentially linear size, $E®=O®$. Later, Hardy and Landau proved that the error cannot be as small as $O(r^{1/2})$; in particular,
$|E®|\neq o\left(r^{1/2}(\log r)^{1/4}\right)$.
The famous conjecture is that the true order is almost square-root size:
$|E®|=O_\varepsilon(r^{1/2+\varepsilon})$
for every $\varepsilon>0$.
The best bound stated in the article is
$|E®|=O(r^{131/208})$
where
$\frac{131}{208}\approx0.6298$.
Thus there remains a significant gap between the known upper exponent $0.6298\ldots$ and the conjectured exponent $1/2$.
Connection with sums of two squares
The problem has a beautiful connection with number theory. Let $r_2(n)$ denote the number of representations of $n$ as a sum of two squares, counting signs and order. Then
$N®=\sum_{n=0}^{r^2}r_2(n)$.
So the geometric problem of counting lattice points becomes a problem about the arithmetic structure of integers and their representations as
$n=m^2+k^2$.
There is also an exact formula involving floor functions, obtained using Jacobi's two-square theorem.
Why the problem is difficult
The main difficulty is that the boundary of a circle does not align with the integer grid. As $r$ changes continuously, $N®$ changes only when the circle crosses a lattice point, whereas $\pi r^2$ changes smoothly. Consequently, the error $E®$ oscillates considerably.
For example, at $r=5$ there are $81$ lattice points, while
$\pi(5)^2=25\pi\approx78.54$,
giving an error of about $2.46$. Small changes in the radius can produce noticeably different errors.
Bigger picture
The Gauss circle problem is a fundamental example of a lattice-point counting problem in analytic number theory. It connects geometry, Fourier analysis, exponential sums, Bessel functions, and the arithmetic of sums of two squares. It also has higher-dimensional analogues involving lattice points inside spheres and other shapes.
In one sentence: The Gauss circle problem asks how closely the number of integer points inside a large circle follows its area $\pi r^2$; although we know the error is substantially smaller than $r$, proving the conjectured near-$r^{1/2}$ error remains an important open problem.
Read the full Wikipedia article
┌────────────────────────────────┐
│ KONSTANTINOS MICHAILIDIS │
└────────────────────────────────┘
│ KONSTANTINOS MICHAILIDIS │
└────────────────────────────────┘