A bird atlas is less a book than a protocol. The following pages describe how 4,128 volunteers, four seasons, and a five-kilometer grid produced the third statewide inventory of New York's breeding birds.
New York is divided into 5,334 atlas blocks, each a square of roughly 5 × 5 kilometers based on the USGS 7.5-minute topographic quadrangle system. Every block has a unique identifier (e.g. 3569D) and an atlas coordinator who watches over it across the five-year survey window.
For each species seen in a block, volunteers record a single letter of evidence. The strongest evidence wins. A singing male is possible. A pair on territory is probable. A nest with eggs, or a fledgling, is confirmed. A block counts a species as breeding on the basis of its best record across the entire five-year window.
Volunteers aimed for a minimum of twenty hours of fieldwork in each priority block, across at least two of the five breeding seasons. Coverage was tracked live inside eBird so coordinators could shift effort into the thin blocks. At the close of 2024, 94.7% of priority blocks had met the effort target.
Comparisons between atlases use both raw occupancy and effort-corrected occupancy. The maps you see on this site are the raw numbers unless the legend says otherwise.
For species with more than 50 observations, we fit block-level occupancy models using a hierarchical Bayesian framework, with detection corrections keyed to effort, season, time of day, and observer skill. Modeled abundance was then derived from Cornell's Status & Trends machinery, cropped to New York's borders. Modeled products are available as GeoTIFF for each species.
We follow the AviList global taxonomy, updated annually. When a species is renamed, the historical records are retained under their original name and cross-referenced to the new one in the database; the web pages display the current common and Latin names. A full synonym table lives in the data appendix.
