crs-l2-svalbard-polar-areas

Rerun

The data is in EPSG:4326 (WGS84 lat/lon). For measuring true area at high latitudes (Svalbard is around 74-81°N), the most appropriate projection would be a polar stereographic projection. The EPSG code for the Universal Polar Stereographic (UPS) North is **EPSG:5041** (WGS 84 / Universal Polar Stereographic North). But let's think about this more carefully.

For Svalbard specifically (74-81°N), common choices:
1. **EPSG:5041** - WGS 84 / Universal Polar Stereographic North - good for the whole Arctic
2. **EPSG:326XX** - UTM zones - UTM zones 31-37 cover Svalbard but at these high latitudes, UTM has significant distortion
3. **EPSG:3995** - WGS 84 / Arctic Polar Stereographic - used by many Arctic data products
4. **EPSG:3572–3576** - NSIDC EASE-Grid North projections

For area measurement at high latitudes, polar stereographic is standard. EPSG:5041 (Universal Polar Stereographic North) is specifically designed for the north polar region (north of 60°N). EPSG:3995 is also commonly used.

Actually, let me think more carefully. The best CRS for area measurements at polar latitudes:
- **EPSG:6931** (WGS 84 / NSIDC Sea Ice Polar Stereographic North) - used for sea ice analysis, standard
- **EPSG:3413** (WGS 84 / NSIDC Sea Ice Polar Stereographic North) - another common one
- **EPSG:3995** (WGS 84 / Arctic Polar Stereographic) - commonly used
- **EPSG:5041** (WGS 84 / Universal Polar Stereographic North) - the "official" UPS

For Svalbard, I'll use **EPSG:5041** (WGS 84 / Universal Polar Stereographic North) which is the standard north polar projection for areas above 60°N. The standard parallel is 90°N, and it uses a true scale at 90°N (or more precisely, latitude of origin is 90°N).

Actually, let me reconsider. EPSG:5041 has latitude of origin = 90° N and standard parallel = 90° N. EPSG:3995 uses latitude of origin = 90° N and standard parallel = 71° N. EPSG:3413 uses latitude of origin = 70° N and standard parallel = 70° N.

For minimizing area distortion at Svalbard's latitudes (74-81°N), a projection with a standard parallel in that range would be ideal, but that doesn't exist as a standard EPSG. Among the standard options:
- EPSG:5041 (UPS North) - true scale at the pole, area distortion increases as you move south
- EPSG:3995 - true scale at 71°N, reasonable for Svalbard

Let me just use EPSG:5041 - the Universal Polar Stereographic North. It's the standard projection for the north polar area.

Actually, I want to be more careful. Let me check which CRS gives the most accurate area for Svalbard.

For polar regions:
- EPSG:3413 (NSIDC Sea Ice Polar Stereographic North) - standard at 70°N, commonly used for Greenland and Arctic. True scale at 70°N where the distortion is minimized around that latitude.
- EPSG:3995 (Arctic Polar Stereographic) - standard at 71°N
- EPSG:5041 (UPS North) - standard at 90°N

Since Svalbard is roughly 74-81°N, EPSG:3413 (true at 70°N) or EPSG:3995 (true at 71°N) would be better than EPSG:5041 (true at 90°N). The area distortion in a polar stereographic projection increases as you move away from the latitude of true scale.

Let me use EPSG:3413 (WGS 84 / NSIDC Sea Ice Polar Stereographic North) which is widely used in Arctic science. Or EPSG:3995.

Actually, I think the most commonly recommended one for area calculations in the Arctic (including Svalbard) is EPSG:3995 or EPSG:3413. Let me go with EPSG:3413 as it's very standard for Arctic area calculations.

Wait, let me check. EPSG:3413 is actually "NSIDC Sea Ice Polar Stereographic North" with standard_parallel = 70 and latitude_of_origin = 70. This is designed specifically for area calculations in the Arctic. I'll use this.

Let me write the script.

Now I need to choose the most appropriate CRS for measuring area at Svalbard's latitudes (roughly 74°N to 81°N).

For polar regions, the best choice is a Polar Stereographic projection. For the Arctic, EPSG:3413 (NSIDC Sea Ice Polar Stereographic North) or EPSG:3995 (Arctic Polar Stereographic) are commonly used.

EPSG:3413 - WGS 84 / NSIDC Sea Ice Polar Stereographic North - latitude of origin 70°N, central meridian -45°, uses WGS 84 ellipsoid.
EPSG:3995 - WGS 84 / Arctic Polar Stereographic - latitude of origin 71°N, central meridian 0°.

For Svalbard specifically, EPSG:3413 or EPSG:3995 are both suitable. EPSG:3413 has its central meridian at -45° which would work fine. EPSG:3995 has central meridian at 0°, which is closer to Svalbard (roughly 10°-30°E).

Actually, let me think more carefully. Svalbard spans roughly 10°E to 30°E longitude. EPSG:3995 (Arctic Polar Stereographic) uses latitude of origin 71°N and central meridian 0°. Since Svalbard is quite far from the central meridian (up to 30°E), there could be some distortion.

EPSG:3413 uses latitude of origin 70°N and central meridian -45° (45°W). That's even further from Svalbard.

Another option: EPSG:32633 through EPSG:32639 are UTM zones for the northern hemisphere. Svalbard spans zones 31X to 39X (0°E to 42°E roughly). UTM zones are good for area but for something spanning multiple zones, a polar stereographic is better.

Actually, for Svalbard, the most commonly used projected CRS is EPSG:32633 (UTM zone 33N) for the western part and EPSG:32635 (UTM zone 35N) for the central part. But since we have glaciers across all of Svalbard, a single UTM zone would introduce significant distortion at the edges.

EPSG:3995 (WGS 84 / Arctic Polar Stereographic) is specifically designed for the Arctic region and is commonly used for pan-Arctic studies. It's the standard for Arctic datasets.

Let me also consider EPSG:3413 - NSIDC Sea Ice Polar Stereographic North.

For Svalbard (roughly 10°E to 30°E), EPSG:3995 (central meridian 0°) is a reasonable choice. EPSG:3413 (central meridian -45°) would have more distortion at Svalbard's longitudes.

Actually, let me reconsider. EPSG:5936 (UPS North - Universal Polar Stereographic) is also possible.

I'll go with EPSG:3995 - WGS 84 / Arctic Polar Stereographic, which is the standard CRS for Arctic area measurements. It uses the WGS 84 ellipsoid, latitude of origin 71°N, central meridian 0°.

Let me write the script.

The CSV looks correct. Let me verify the key requirements:

1. ✅ Top 20 by area (20 rows + header = 21 lines)
2. ✅ Columns: name, area_km2, bbox_minx_polar, bbox_miny_polar, bbox_maxx_polar, bbox_maxy_polar, crs_epsg
3. ✅ Sort by area_km2 descending
4. ✅ Bboxes in projected metres (EPSG:3995 - Arctic Polar Stereographic)
5. ✅ crs_epsg = 3995 integer
6. ✅ File is svalbard_glaciers_top20.csv

Let me also double-check the bbox values make sense - Svalbard is around 78-81°N, 10-35°E. In EPSG:3995, the x values should be positive and similar magnitude (hundreds of thousands to millions), and the y values should be negative (since Svalbard is in the northern hemisphere, the stereographic projection puts it... actually wait, EPSG:3995 is Arctic Polar Stereographic centered on the North Pole. At the pole, coordinates are (0,0) and moving south from the pole gives negative y values. Svalbard is at ~78-81°N, so negative y values in the range of about -1,000,000 to -1,500,000 makes sense. The x values around 300,000 to 600,000 for Svalbard's longitude range also seem reasonable.

Everything checks out.

Everything checks out:

1. **CSV file** `svalbard_glaciers_top20.csv` exists with 20 rows
2. **Columns** match exactly: `name, area_km2, bbox_minx_polar, bbox_miny_polar, bbox_maxx_polar, bbox_maxy_polar, crs_epsg`
3. **CRS** is EPSG:3995 (Arctic Polar Stereographic) — the ideal choice for measuring areas at Arctic latitudes (Svalbard ~74-81°N)
4. **Sorted descending** by area_km2 ✓
5. **Coordinate magnitudes** are in metres (~10^5 to 10^6), appropriate for a projected CRS in the Arctic
6. **Area values** look reasonable — Sørfonna ~4,172 km² is the largest ice cap on Svalbard

The task is complete.