研究目的
To develop and present cost-effective approaches that rely on observations extracted from a wide range of readily available remote sensing datasets, and procedures that could be applied to vulnerable coastal areas and inland bodies of water worldwide for algal bloom modeling.
研究成果
The developed methodologies are advantageous as they rely on readily available global remote sensing datasets, are cost-effective, identify factors controlling algal bloom occurrences, and can potentially assist in the assessment and understanding of the complexities associated with algal bloom development worldwide. The hybrid model improved the model performance significantly, indicating the effectiveness of the two-step process in predicting algal bloom probability.
研究不足
The main constraints are related to the spatial and temporal resolution of the remote sensing digital datasets used. Algal blooms smaller than the picture elements of digital products or those that occur and vanish within the acquisition period of two consecutive satellite images will not be detected. Adverse weather conditions can degrade the quality of remote sensing datasets.
1：Experimental Design and Method Selection:
The study adopted a fivefold methodology involving the generation of a database incorporating relevant coregistered remote sensing datasets and derived products, compilation of an inventory of known locations of algal blooms, investigation of factors controlling algal bloom occurrences, construction and validation of MR and ANN models, and adoption of a two-step modeling approach (temporal model followed by a spatial model) to improve prediction accuracy.
2：Sample Selection and Data Sources:
The study utilized remote sensing datasets from MODIS, TRMM, and QuikSCAT, along with field data provided by the Kuwait Institute for Scientific Research (KISR).
3：List of Experimental Equipment and Materials:
Software SeaDAS
4：4, ENVI 8, and ArcGIS 1 were used for image processing and GIS analyses. Experimental Procedures and Operational Workflow:
The methodology involved data preprocessing, model construction using training data subsets, and model evaluation using testing data subsets.
5：Data Analysis Methods:
The study employed ROC tests for model validation and comparison, and used multivariate regression and artificial neural network techniques for data analysis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
