A new equal-area isolatitudinal grid on a spherical surface

The paper proposes SREAG, a new method for creating a spherical rectangular equal-area grid that divides a sphere into latitudinal rings of near-constant width and further splits them into equal-area cells, offering superior uniformity, resolution flexibility, and ease of use compared to existing equal-area pixelization techniques.

The Gist

Imagine you are trying to wrap a giant, perfectly round beach ball (the Earth or the sky) with a sheet of graph paper. Your goal is to cut that paper into tiny, equal-sized squares so you can count things on the ball—like stars, radio signals, or weather patterns—without any part of the ball being "squished" or "stretched" more than another.

This is the problem that Zinovy Malkin solves in this paper. He introduces a new way to slice up a sphere called SREAG (Spherical Rectangular Equal-Area Grid).

Here is the breakdown of the problem and his solution, using simple analogies:

The Problem: The "Orange Peel" Dilemma

If you try to peel an orange and lay the skin flat on a table, it rips and stretches. You can't make a perfect flat map of a round ball without distorting something.

In the past, scientists used a few different methods to slice up the sky, but they all had flaws:

1. The "Standard Map" Method: Imagine drawing a grid on a flat map of the world. If you wrap that map around a globe, the squares near the equator look like squares, but near the North Pole, they get squished into tiny, skinny triangles. The area isn't equal.
2. The "Lambert" Method: This tries to keep the area equal, but the shapes get weird and distorted, like a funhouse mirror.
3. The "HEALPix" Method: This is currently the most popular method in astronomy. It uses diamond-shaped tiles. It's great, but the "rings" of latitude (horizontal bands) aren't all the same height. Some are tall, some are short, which makes it hard to compare data horizontally.

The Goal: Scientists wanted a grid that had three perfect traits at once:

• Every single tile has the exact same area (so a count of 10 stars means the same thing everywhere).
• The tiles are rectangular (aligned with North-South and East-West), making them easy to use with standard maps.
• The horizontal "rings" are all the same height (uniform width).

Until now, nobody thought you could have all three at once.

The Solution: The "Adjustable Belt" Strategy

Malkin's new method, SREAG, is like a clever way of adjusting belts on a stack of hula hoops.

Step 1: The Rough Cut
First, he cuts the sphere into horizontal rings (like slicing a loaf of bread, but the slices are rings). He makes sure the rings are all roughly the same height. Then, he cuts each ring into vertical slices (like cutting a pizza).

• The Trick: Near the equator, the ring is wide, so he makes the slices wide. Near the poles, the ring is narrow, so he makes the slices narrow. This ensures the pieces are roughly square-shaped near the middle.

Step 2: The Fine-Tuning (The Magic)
Here is where the math gets clever. He realizes that if he keeps the rings exactly the same height, the areas won't be perfectly equal. So, he slightly shifts the boundaries of the rings.

• Imagine the rings are made of stretchy rubber. He stretches the rings near the poles slightly and squeezes the rings near the equator slightly.
• He does this just enough so that every single tile (no matter where it is on the ball) covers the exact same amount of surface area.

Why is this a Big Deal?

Think of it like organizing a library.

• Old Methods: Some shelves were huge (holding 100 books) and some were tiny (holding 1 book). If you wanted to know how many books were in the "North" section, you had to do complicated math to figure out if you were comparing apples to oranges.
• The New SREAG Method: Every shelf is exactly the same size. Every box holds exactly 10 books. It's a perfect grid.

The Benefits:

1. Fairness: Because every tile is the same size, counting stars or radio sources is fair. You aren't accidentally counting more stars just because your grid tile is bigger.
2. Simplicity: The tiles are perfect rectangles (North-South, East-West). Astronomers and geographers are used to thinking in Latitude and Longitude. This grid fits their brains perfectly.
3. Flexibility: You can make the grid as detailed as you want. You can have 20 big tiles or 10 million tiny tiles. The math works for any number.
4. Uniformity: The horizontal bands are almost perfectly the same height, which makes comparing data across the sky much easier than with previous methods.

The Bottom Line

Malkin has invented a new "graph paper" for the universe. It solves a centuries-old geometry puzzle by slightly bending the rules of how we cut the rings, ensuring that every piece of the puzzle is the same size and shape. This makes it much easier for scientists to analyze data from space, whether they are looking at the entire sky or just a tiny patch of it.

It's a simple, elegant solution that turns a messy, distorted problem into a clean, organized grid.

Technical Summary

1. Problem Statement

The subdivision (pixelization) of a spherical surface into equal-area cells is a fundamental requirement in astronomy and geodesy for tasks such as data averaging, error mitigation, optimizing spherical function computations, and cross-analyzing differently sampled datasets.

Existing methods suffer from specific limitations regarding the trade-off between equal cell area and uniform latitudinal ring width:

• Equirectangular Projection: Cells have uniform longitude/latitude spans, but cell area decreases significantly toward the poles.
• Lambert Azimuthal Equal-Area Projection: Provides equal-area cells on a plane, but when projected onto a sphere, the latitudinal ring widths become non-uniform.
• HEALPix: Widely used in astronomy, it provides equal-area rhombus-like cells. However, it lacks true latitudinal bands (rings), and the "rings" formed by cell centers have highly non-uniform widths compared to a uniform distribution.
• Previous Iterative Methods (Malkin 2016a): Attempted to adjust boundaries iteratively but resulted in poor uniformity of cell area and were computationally expensive for large grids.

The Goal: To construct a grid that satisfies five specific criteria:

1. Rectangular cells with boundaries oriented along latitude and longitude circles.
2. Strictly uniform cell area across the entire sphere.
3. Uniform width of latitudinal rings.
4. Near-square cells in equatorial rings.
5. Simple computational realization (forward and inverse mapping).

The author notes that a solution satisfying all criteria perfectly (specifically simultaneous equal area and equal ring width) likely does not exist mathematically. The paper proposes a method that prioritizes equal cell area while maximizing the uniformity of ring widths.

2. Methodology: SREAG (Spherical Rectangular Equal-Area Grid)

The proposed SREAG method employs a two-step construction strategy:

Step 1: Initial Grid Construction

1. Latitudinal Division: The sphere is initially divided into $N_{ring}$ latitudinal bands of constant width ($dB$).
2. Cell Span Calculation: Each ring is divided into cells. The longitudinal span ($dL_i$) for ring $i$ is calculated based on the central latitude ($b_{0,i}$) of that ring:
   $$dL_i = dB \cdot \sec(b_{0,i})$$
   This ensures that cells in the equatorial rings are nearly square.
3. Integer Rounding: The number of cells per ring is calculated as $360^\circ / dL_i$ and rounded to the nearest integer. This creates an "Initial Grid" that satisfies the rectangular shape and near-uniform ring width but fails the strict equal-area requirement.

Step 2: Boundary Adjustment (The Core Innovation)

To achieve strict equal-area cells, the latitudinal boundaries are adjusted while keeping the longitudinal span ($dL_i$) fixed to the values calculated in the initial grid.

• Target Area ($A$): The area of a single cell is defined as the total surface area of the unit sphere ($4\pi$) divided by the total number of cells ($N_{cell}$).
• Iterative Calculation: Starting from the North Pole ($b_u = \pi/2$), the lower boundary ($b_l$) of each ring is computed using the area formula for a spherical zone:
  $$A = dL_i (\sin b_u - \sin b_l)$$
  Rearranging for the lower boundary:
  $$b_l = \arcsin(\sin b_u - A/dL_i)$$
• Symmetry: The process iterates down to the equator. The boundaries for the Southern Hemisphere are generated by mirroring the Northern Hemisphere boundaries with a negative sign.

This procedure ensures that every cell has exactly the same area $A$, while the latitudinal ring widths deviate only minimally from the ideal uniform width.

3. Key Contributions

• New Algorithm: Introduction of the SREAG method, which is mathematically rigorous, simple to implement, and allows for the construction of grids with an arbitrary number of rings.
• Optimization of Trade-offs: The method provides the best known compromise between strictly equal cell areas and near-uniform latitudinal ring widths.
• Rectangular Geometry: Unlike HEALPix (rhombus) or other projections, SREAG maintains a rectangular grid structure aligned with the celestial coordinate system (Right Ascension/Declination), facilitating easier data visualization and interpretation.
• Scalability: The method is not limited to specific resolution steps (powers of 2, as in HEALPix). It supports a theoretically unlimited range of resolutions, constrained only by machine precision.
• Software Availability: The author provides Fortran routines for implementation.

4. Results and Analysis

The paper presents a comparative analysis of SREAG against the Lambert projection and HEALPix (various $N_{side}$ values).

• Uniformity of Ring Widths:
  • SREAG: The deviation between the actual central latitude of a ring and the nominal (uniform) central latitude is extremely small. For grids with $N_{ring} > 24$, this deviation decreases monotonically.
  • Comparison: In Table 2, for a grid with ~48 cells, HEALPix ($N_{side}=2$) shows deviations of up to -10.70°, whereas SREAG (46 cells) shows a maximum deviation of 0.93°. Even for larger grids (e.g., ~192 cells), HEALPix deviations remain around -5° to -6°, while SREAG deviations are negligible (<0.4°).
• Resolution Flexibility: Figure 2 demonstrates a nearly linear log-log relationship between the number of rings and the number of cells, offering a much finer granularity of resolution choices compared to the discrete steps of HEALPix.
• Cell Shape: The method successfully produces near-square cells in equatorial regions, as intended by the initial $\sec(b_0)$ scaling.

5. Significance and Applications

The SREAG method addresses a critical gap in spherical data analysis by offering a grid that is:

1. Astronomically Intuitive: The rectangular, isolatitudinal nature aligns perfectly with standard astronomical coordinate systems, making binned data easy to visualize and interpret.
2. Computationally Efficient: The direct calculation of boundaries avoids the time-consuming iterative adjustments required by previous methods.
3. Versatile: It is suitable for both large-scale structure analysis and tiny-scale investigations.

Practical Applications Cited:

• Resampling: Used to convert non-uniform distributions of radio stars into uniform samples.
• Source Distribution Analysis: Applied to analyze the distribution of giant radio sources and sources in the International Celestial Reference Frame (ICRF3).
• Systematic Error Evaluation: Useful for evaluating systematic errors in source position catalogs.

In conclusion, SREAG represents a significant advancement in spherical pixelization, prioritizing the critical requirement of equal cell area while maintaining a geometry that is highly favorable for astronomical and geodetic applications.