Document Type : Original Article
Drought is a prevalent environmental hazard across diverse climates, resulting in significant economic and social consequences. Agricultural drought is influenced by various meteorological factors and plant characteristics such as plant species, growth period duration, and phenology traits. Consequently, in regions reliant on rainfed agriculture, such as arid and semi-arid areas, conducting drought risk assessments is crucial for evaluating agricultural output, public health, and food security. Given Iran's predominantly dry and semi-arid climate, monitoring and analyzing drought occurrences are imperative, prompting numerous research endeavors in this domain. Several indices have been devised to streamline trend analysis and statistical forecasting to facilitate drought analysis, with notable examples including the Standardized Precipitation Index (SPI) and the Reconnaissance Drought Index (RDI). Notably, the enhanced version of the RDI index, known as the effective RDI (eRDI) index, has garnered attention for its superior performance in various sources. This study compares the RDI and eRDI indices over different periods to ascertain the yield trends of rainfed wheat in North Khorasan province, representing a novel contribution to the field. Incorporating effective precipitation in the eRDI index necessitates the evaluation of diverse methodologies to determine and inform the index equations. Given North Khorasan's unique climatic conditions and significant role in national wheat production, the province holds particular importance, yet there remains a scarcity of agricultural drought studies in the region, underscoring the critical nature of this issue in the area.
In order to assess agricultural drought conditions, monthly rainfall and temperature data from six cities within a province were analyzed over a consistent statistical timeframe. The evaluation of index data involved utilizing information on wheat cultivation performance across various cities spanning 18 years from 1989 to 2021, sourced from governmental statistical reports specific to North Khorasan province; different methods of calculating effective precipitation in the eRDI model, including those by FAO, USDA, USBR, and USDAsim, were compared. The efficacy of RDI and eRDI indices in reflecting drought impacts was assessed by correlating wheat yield with wheat production over time. Standardized performance metrics for each city were employed to normalize the data to account for variations in crop performance due to factors such as soil type and management practices. Statistical analyses, the Kolmogorov-Smirnov test to assess normality, appropriate correlation coefficient calculations, and comparison of averages, were conducted to compare data across different stations. Trend analysis of the selected index was performed using non-parametric Mann-Kendall tests and Sen's estimator.
Monthly temperature data were utilized to compute potential evaporation employing the Thornthwaite technique. The findings revealed that the Jajarm station, characterized by arid climatic conditions, exhibited the highest average evaporation rate within the province, while the Faruj station, experiencing semi-arid climatic conditions, displayed the lowest evaporation rate. Furthermore, the aridity index, denoting the ratio of average precipitation to average annual evaporation and transpiration, was determined for various stations. The USDAsim method was selected among four methods to estimate effective precipitation. Evaluation of the root mean square error between the two indices indicated that their disparity was minimal in the arid Jajarm station, escalating as the region's humidity increased. Pearson's correlation coefficient values every month across all stations demonstrated a strong correlation between the two indices. Additionally, a t-test was conducted to compare the monthly averages of the two indices, revealing no significant distinction at the 5% level across any of the stations. Analysis of the correlation coefficient between the indices and the annual wheat yield indicated that specific periods were influential in different stations. For instance, in Shirvan and Faruj, the period from April to June was significant, while in Mane and Samalghan, the period from November to January was crucial. In Bojnord, the period from November to May was influential, and in Jajarm, the period from October to June was significant. Trend analysis unveiled a negative slope solely for the Bojnord and Esfarayen stations, with statistical significance observed only at the Bojnord station using the Mann-Kendall test at the 5% level. Conversely, positive trends, albeit insignificant, were noted in other stations. Notably, the most pronounced decreasing and increasing trends were associated with the Bojnord and Faruj stations, respectively. Graphical representation of the Mann-Kendall test illustrated a declining trend in Bojnord, Esfarayen, Mane, and Samalghan, no discernible trend in Shirvan, and a non-significant upward trend in Faruj.
The significance of this study lies in the unique phenology of wheat and its heightened vulnerability to moisture deficiency during specific growth stages. Therefore, it is imperative to establish a distinct reference time frame for each city based on climate conditions. The eRDI accurately incorporates the concept of usable plant precipitation in its calculations. Analysis of the eRDI method's periods reveals that each city exhibits a unique performance period based on its climatic and geographical characteristics, highlighting the necessity to reassess the utilization of specific time frames in comparative studies across different cities. Trend analysis results indicate the effectiveness of the Mann-Kendall test and sen's estimator in identifying yield trends, such as the declining trend in Bojnord and Esfarayen and the increasing trend in other locations. While a longer-term dataset is required for a more accurate trend diagnosis, these findings can inform analysis and planning efforts in food security, particularly in areas under wheat cultivation.
The authors declare no conflict of interest.
