The optimal time-scale of Standardized Precipitation Index for early identifying summer maize drought in the Huang-Huai-Hai region, China

The Huang-Huai-Hai region (HHH). The optimal time-scales of Standardized Precipitation Index (SPI) for early drought identification was determined by using meteorological data and agrometeorological disaster observation records of summer maize in HHH in this study. Two performance measures were used, one is accuracy, defined as the proportion of the drought samples identified by SPI in the total samples, and the other is sensitivity, defined as the ratio of the number of early recognized samples to the number of identified samples. The SPI can be calculated with only precipitation at any time-scales, mainly including 3-, 6-, 12-, or 24-month periods in previous studies. However, the growth period of maize is only 3–4 months, and the SPI with more than 3-month time-scales may not be applicable for maize drought identification. Additionally, meteorological drought is considered to be the origin for agricultural drought. It is important to identify meteorological drought signals in advance by using applicable time-scale SPI to avoid crop yield reduction. Results showed that 10-day SPI (SPI10) was the optimal index for drought identification with the highest accuracy (95.8 %) and sensitivity (77.72 %). Additionally, most of the drought samples identified by SPI10 were 0–16 days ahead of the observed onset date. The findings can provide theoretical guidance for early warning and near real-time monitoring for summer maize drought disasters in HHH. [Display omitted] •10-day is the optimal time-scale for SPI to early identify drought of summer maize in the HHH region.•SPI10 has a higher accuracy (95.8 %) and sensitivity (77.71 %) to identify drought of summer maize in the HHH region.•The droughts could be identified 0–16 days ahead of the actual onset date by SPI10.•Typical drought processes validated the applicability of SPI10.