In the last few decades, harmful algal blooms (HABs, also known as “red tides”) have become one of the most detrimental natural phenomena in Florida’s coastal areas. Karenia brevis produces toxins that have harmful effects on humans, fisheries, and ecosystems. In this study, we developed and compared the efficiency of state-of-the-art machine learning models (e.g., XGBoost, Random Forest, and Support Vector Machine) in predicting the occurrence of HABs. In the proposed models the K. brevis abundance is used as the target, and 10 level-02 ocean color products extracted from daily archival MODIS satellite data are used as controlling factors. The adopted approach addresses two main shortcomings of earlier models: (1) the paucity of satellite data due to cloudy scenes and (2) the lag time between the period at which a variable reaches its highest correlation with the target and the time the bloom occurs. Eleven spatio-temporal models were generated, each from 3 consecutive day satellite datasets, with a forecasting span from 1 to 11 days. The 3-day models addressed the potential variations in lag time for some of the temporal variables. One or more of the generated 11 models could be used to predict HAB occurrences depending on availability of the cloud-free consecutive days. Findings indicate that XGBoost outperformed the other methods, and the forecasting models of 5–9 days achieved the best results. The most reliable model can forecast eight days ahead of time with balanced overall accuracy, Kappa coefficient, F-Score, and AUC of 96%, 0.93, 0.97, and 0.98 respectively. The euphotic depth, sea surface temperature, and chlorophyll-a are always among the most significant controlling factors. The proposed models could potentially be used to develop an “early warning system” for HABs in southwest Florida.

No abstract

No abstract

<p>Climate change increases the probability of drought occurrence in many parts of the United States and worldwide. Aquifer response to these drought events vary in space and time. This project seeks to understand the response of aquifers to drought events by quantifying the lag time between meteorological droughts and groundwater droughts using the Standardized Precipitation and Evapotranspiration Index (SPEI) and Gravity Recovery and Climate Experiment (GRACE) derived groundwater storage anomalies. Ten major aquifer systems in the continental United States were selected for analysis: Columbia Plateau, Arizona Alluvial, Snake River Basin, Upper Colorado, Pennsylvanian, Mississippi Embayment, Texas Gulf Coast, Edwards-Trinity Plateau, Floridian, Central California, and the High Plains Aquifer Systems. Groundwater storage anomaly data was derived from GRACE total water storage anomaly data by removing all other hydrologic components using the Global Land Data Assimilation System’s (GLDAS) Community Land Surface Model (CLM) of 1.0-degree spatial resolution monthly datasets. Timeseries on monthly intervals for both the derived groundwater storage and SPEI were created for the period of April 2002 to June 2021. Each selected aquifer system had a meteorological drought occur at least three times during the study period, with a maximum occurrence of fifteen in central California. There is a temporal gap in between the original GRACE mission and the launch of GRACE-Follow on (GRACE-FO) from June 2017 to June 2018, five of the ten selected aquifers had meteorological droughts occur in this gap, which have been excluded. Preliminary results indicate that the lag time between the start of the two types of droughts for these aquifer systems is between zero and one month, while the lag time between the end of these types of droughts is more widely varied, between zero and eight months. As these results are varied, contextualizing them with more in-depth looks at the aquifer system characteristics is important and is the next step in furthering our understanding of aquifer responses to the increasing number of probable drought events.</p>