Bird collisions with offshore wind structures are difficult to measure. Assessing collision rates at other human structures, like tall buildings and land-based wind facilities, is typically done by repeatedly counting carcasses underneath the structures. This is not possible in the ocean as bird carcasses sink, float away, or are scavenged. Thus, direct observations or remote monitoring technologies such as cameras and thermal sensors must be used. Monitoring bird collisions in the offshore environment (whether via direct observation or remote monitoring) presents significant logistical and technical difficulties, particularly for smaller-bodied birds and species that migrate at night. Thus, while there are several key studies that have used a combination of human observers, camera systems, and radar to count collisions at offshore wind facilities, this type of data collection at scale is expensive and technically demanding. Prior to construction, collision risk models (CRMs) are used during the permitting process to estimate the likelihood of bird collisions at a given wind facility site. Such models use wind facility characteristics (e.g., turbine height) and species-specific parameters (e.g., distribution, avoidance behavior, morphology) to predict potential collision rates. However, there are limited data available to validate predictions from these models, and more field research is needed to improve confidence in model outputs. In the face of limited data, however, CRMs can help inform risk assessments and possible mitigation strategies. For more detailed information and scientific citations, please see the full FAQ document linked below.