All work
Work
Each project comes with its data and its verification. Filter by topic, or open a card.
Texas wind power: how much room does the official forecast leave?
Day-ahead error is 6.24%, level with a perfect-forecast ceiling; the high bias is curtailment. Driven by Pangu-Weather or GraphCast operational forecasts, the same model falls clearly behind the official forecast.
Where the dust came from: the March 2023 East Asian dust storm
FLEXPART-WRF splits downstream PM₁₀ between two source regions; CALIOP overpasses show the model lofts Taklimakan dust about 0.9 km too low.
Radar extrapolation vs. radar–NWP blending: who wins?
Three U.S. cases: blending leads at every lead time in a supercell but loses at +1 h in stratiform rain. The evaluation became nowcast-bench, a package that reports sample sizes with every score.
Radar echo-arrival warning system
Pushes a phone alert when echoes approach a county. Six single-site radars plus a regional mosaic, registered to county boundaries, with three rules to suppress clutter alarms.
Dust exposure of the northwest China solar corridor
Siting structure of solar farms and trends in their dust exposure; no estimate of generation loss.
Code is linked where public; the rest is available on request.