Recently, an in-situ supplementary irrigation device (SID) has been developed, which can significantly improve the survival rate of saplings by collecting rainwater for irrigation using ceramic emitters, yet the suitable structure parameters (rated area, rated capacity, and rated flow) of in-situ SID still remain unclear due to the difference of precipitation at various sites of dryland. To address that, the present study conducted a novel algorithm that could directly determine the suitable structure parameters of the in-situ SID according to data of the annual mean precipitation. In order to develop this algorithm, we firstly developed a daily water balance model based on daily precipitation data from 96 sites in dryland, and then determined the suitable structure parameters according to the highest operating reliability. After that, the relationship between the optimal structure parameters and the predictors, including geographical and annual meteorological factors, was developed using multiple linear regression. Hence, the suitable structure parameters of various sites can be quickly determined. Moreover, a validation experiment was carried out at Tibet Lhasa. The results showed that the survival rate of saplings was 83.6% under in-situ SID with suitable structure parameters using the developed algorithm, which was 36.7% higher than that of traditional planting methods. Finally, the present provided a convenient table according to the established algorithm, which showed the suitable structure parameters of in-situ SID for extremely arid, arid, and semiarid regions, as well as the corresponding pioneer sapling species.