LAUSR

Two disadvantages of the entropy-weighting model (EWM) in drought-risk assessment are presented through two typical examples in this paper. (1) For distortion in the normalization process, entropy defined by EWM cannot represent the indicator’s dipartite degree correctly when too many zero values exist in the observation data. (2) Given that EWM neglects the indicator’s practical significance in drought-risk assessment, the indicator’s dipartite degree cannot correctly represent its importance when observation data are concentrated in the worst category. These two problems lead to unjustified drought-risk assessment results. Therefore, the features of observation data should be checked before weighting. If the indicator’s observation values are concentrated in the worst domain or numerous zero values exist, then EWM should be applied cautiously.