Search for: All records

Tropical cyclone (TC) models indicate that continued planet warming will likely increase the global proportion of powerful TCs (specifically Categories 4 and 5 hurricanes), increasingly jeopardizing low-lying coastal communities and resources such as the Pelican Cays, Belize. The combination of increased coastal development and continued relative sea-level rise puts these communities at even higher risk of damage from TCs. The short TC observational record for the western Caribbean hampers the extensive study of TC activity on centennial timescales, which hinders our ability to fully understand past TC climatology and improve the accuracy of TC models. To better assess TC risk, paleotempestological studies are necessary to put future scenarios in perspective. Here, we present a high-resolution reconstruction of coarser-grained sediment deposits associated with TC (predominately ≥ Category 2 hurricanes) passages over the past 1200 years from Elbow and Lagoon Cays, two coral reef-bounded lagoons at the northern and southern end of the Pelican Cays; the most southern Belizean paleotempestological site to date. Coincident timing of historic storms with statistically significant coarser-grained deposits within cay lagoon sediment cores allows us to determine which historic TCs likely generated event layers (tempestites) archived in the sediment record. Our compilation frequency analysis indicates one active interval (above-normal TC activity) from 1740-1950 CE and one quiet interval (below-normal TC activity) from 850-1018 CE. The active and quiet intervals in the Pelican Cays composite record are anticorrelated with those from nearby and re-analyzed TC records to the north, including the Great Blue Hole (∼100 km north) and the Northeast Yucatan (∼380 km northwest). This site-specific anticorrelation in TC activity along the western Caribbean indicates that we cannot rely on any one single TC record to represent regional TC activity. However, we cannot discount that these anticorrelated periods between the western Caribbean sites are due to randomness. To confirm that the anticorrelation in TC activity among sites from the western Caribbean is indeed a function of climate change and not randomness, an integration of more records and TC model simulations over the past millennium is necessary to assess the significance of centennial-scale variability in TC activity recorded in reconstructions from the western Caribbean. 

Abstract Event‐based paleohurricane reconstructions of the last millennium indicate dramatic changes in the frequency of landfalling hurricanes on centennial timescales. It is difficult to assess whether the variability captured in these paleorecords is related to changing climate or randomness. We assess whether centennial‐scale active and quiet intervals of intense hurricane activity occur in a set of synthetic storms run with boundary conditions from an earth system model simulation of the last millennium. We generate 1,000 pseudo sedimentary records for a site on South Andros Island using a Poisson random draw from this synthetic storm data set. We find that any single pseudo sedimentary record contains active and quiet intervals of hurricane activity. The 1,000‐record ensemble average, which reflects the common signal of climate variability, does not. This suggests that the record of paleohurricane activity from The Bahamas reflects variability in hurricane frequency dominated by randomness and not variability in the climatic conditions. 

Abstract Classifying images using supervised machine learning (ML) relies on labeled training data—classes or text descriptions, for example, associated with each image. Data‐driven models are only as good as the data used for training, and this points to the importance of high‐quality labeled data for developing a ML model that has predictive skill. Labeling data is typically a time‐consuming, manual process. Here, we investigate the process of labeling data, with a specific focus on coastal aerial imagery captured in the wake of hurricanes that affected the Atlantic and Gulf Coasts of the United States. The imagery data set is a rich observational record of storm impacts and coastal change, but the imagery requires labeling to render that information accessible. We created an online interface that served labelers a stream of images and a fixed set of questions. A total of 1,600 images were labeled by at least two or as many as seven coastal scientists. We used the resulting data set to investigate interrater agreement: the extent to which labelers labeled each image similarly. Interrater agreement scores, assessed with percent agreement and Krippendorff's alpha, are higher when the questions posed to labelers are relatively simple, when the labelers are provided with a user manual, and when images are smaller. Experiments in interrater agreement point toward the benefit of multiple labelers for understanding the uncertainty in labeling data for machine learning research.