Netcdf Viewer | DIRECT • 2027 |

She clicked a point north of Svalbard. A line of white text appeared in the air: -1.8°C . She dragged her finger across a touchpad that wasn't there—the time slider. The weeks melted forward. March. April. She watched the ice edge retreat like a shy animal, fracturing into the Fram Strait.

Dr. Elara Vance rubbed her eyes. The terminal window glowed with lines of text, a lifeless summary of five years of Arctic ice dynamics. The data was all there—temperature, salinity, pressure, ice thickness—neatly packed into a single, stubborn NetCDF file named arctic_basin_2024.nc . netcdf viewer

Søk would sniff the file. It would find the dimensions—time, latitude, longitude, maybe depth. Then, it would guess. Is tos sea surface temperature? Is siconc sea ice concentration? It would map the first 3D variable to space and the first time dimension to an invisible slider. She clicked a point north of Svalbard

Her colleague, Ben, had tried to walk her through Python scripts again. xarray , matplotlib , cartopy —she could coax out a static plot, a slice through time. But she couldn’t see it. She couldn’t feel the Beaufort Gyre turning or watch the flaw leads crack open. The command line was a wall between her and the story the data was trying to tell. The weeks melted forward

She pushed a final commit that afternoon, adding a subtitle to the project’s README:

For the first time, she saw the whorl . A massive, slow-motion cyclone of ice in the Beaufort Sea, a feature her scripts had reduced to a single standard deviation in a statistics report. She gasped.

The principle was simple. Most NetCDF viewers were either glorified spreadsheet browsers or required a supercomputer. Elara wanted something that felt like holding a snow globe. She wrote the core in Rust for speed, using wgpu for graphics. The interface had no menus, just a void and a prompt.