Abstract: Harmful algal bloom (HAB) occurs frequently and causes serious damage. To study the earlywarning and prediction technology of HAB is of significance for the early-warning and prediction, ecological control, and disaster prevention and mitigation of HAB. This paper reviewed the research progress in the early-warning and prediction technologies of HAB, including transport prediction, specific factors critical value prediction, datadriven model, and ecological math model, and evaluated the advantages and disadvantages of these four types of technologies. Some new ideas were brought forward about the prediction of cyanobacterial growth rate based on cell characteristics, and the earlywarning of cyanobacterial bloom based on algal community characteristics.

Key words: harmful algal bloom, early-warning and prediction, cell characteristics, community characteristics, vegetation coverage, landscape pattern, topographic factors, types of frozen ground.