Drought prediction and sustainable development of the ecological environment

Assessment of drought conditions and prediction by machine learning algorithms using Standardized Precipitation Index and Standardized Water-Level Index (case study: Yazd province, Iran)

Drought as a natural phenomenon has always been a serious threat to regions with hot and dry climates. One of the major effects of drought is the drop in groundwater level. This paper focused on the SPI (Standardized Precipitation Index) and SWI (Standardized Water-Level Index) to assess meteorological and hydrological drought, respectively. In the first part, we used different time frames of SPI (3, 6, 12, and 24 months) to investigate drought in Yazd, a dry province in the center of Iran for 29 years (1990-2018). Then, in the second part, the relationship between SPI and SWI was investigated in the three aquifers of Yazd by some rain gauge stations and the closest observation wells to them. In addition to using SPI and SWI, we also used different machine learning (ML) algorithms to predict drought conditions including linear model and six non-linear models of K_Nearest_Neighbors, Gradient_Boosting, Decision_Tree, XGBoost, Random_Forest, and Neural_Net. To evaluate the accuracy of the mentioned models, three statistical indicators including Score, RMSE, and MAE were used. Based on the results of the first part, Yazd province has changed from mild wet to mild drought in terms of meteorological drought (the amount of rainfall according to SPI), and this condition can worsen due to climate change. The models used in ML showed that SPI-6 (score ave = 0.977), SPI-3 (score ave = 0.936), SPI-24 (score ave = 0.571), and SPI-12 (score ave = 0.413) indices had the highest accuracy, respectively. The models of Neural_Net (score ave = 0.964-RMSE ave = 0.020-MAE ave = 0.077) and Gradient_Boosting (score ave = 0.551-RMSE ave = 0.124-MAE ave = 0.248) had the highest and lowest accuracy in prediction of the SPI in all four-time scales. Based on the results of the second part, about the SWI, Random_Forest model (score = 0.929-RMSE = 0.052-MAE = 0.150) and model of Neural_Net (score = 0.755-RMSE = 0.235-MAE = 0.456) had the highest and lowest accuracy, respectively. Also, hydrological drought (reduction of the groundwater level) of the region has been much more severe, and according to the low correlation coefficient of average SPI and SWI (R2 = 0.14), we found that the uncontrolled pumping wells, as a main factor than a shortage of rainfall, have aggravated the hydrological drought, and this region is at risk of becoming a more arid region in the future.

An innovative bivariate approach to detect joint temporal trends in environmental conditions: Application to large French rivers and diadromous fish

Most key life-events of organisms are synchronized by complex interactions of several environmental cues to ensure optimal survival and growth of individuals and their offspring. However, global change is known to affect multiple components of ecosystems and cues at the same time. Therefore, detecting joint trends in covariate time series is a crucial challenge in global change ecology that has rarely been addressed so far. In this context, we designed an innovative combination of kernel density estimations and Mann-Kendall trend tests to detect joint temporal trends in a pair of environmental variables. This methodological framework was tested on >30 years (1976-2019) of water temperature and discharge data for 6 large French rivers (the Garonne, Dordogne, Rhône, Rhine, Loire and Vienne rivers). The implications of such trends in both temperature and discharge for diadromous species key life-cycle processes were then explored by checking if significant bivariate environmental changes occurred during seasons of upstream and downstream migration, and reproductive activities. Results were contrasted between rivers and seasons: many rivers displayed an increase in the number of days with high water temperature and low river discharge, but local discharge regulation measures could have mitigated the trend in discharge. Our findings showed that species migrating or spawning in spring were likely to be strongly impacted by the new environmental conditions in the Garonne, Loire and Rhône rivers, given the marked changes in water temperature and discharge associations detected by our new method. Conditions experienced by fall-running and spawning species have been strongly affected in all the rivers studied. This innovative methodology was implemented in a new R package, ChocR, for application to other environments and ecosystems.

Responses of meteorological drought-hydrological drought propagation to watershed scales in the upper Huaihe River basin, China

Spatial scale is a crucial factor in drought propagation and characterization. To investigate the response behavior of the meteorological drought-hydrological drought propagation to different watershed scales, three nested subbasins in the upper Huaihe River basin were selected as case study sites. The Standardized Precipitation Index and the Standardized Streamflow Index were adopted to characterize the meteorological drought and hydrological drought respectively. The relationship between meteorological drought and hydrological drought was then examined by wavelet coherency analysis and linear/nonlinear regression models. Results showed that (1) linear regression model captured drought propagation best in all subbasins, and the correlation strengthened as the watershed area increased with drought duration being highest correlated among the five drought characteristics; (2) with the watershed area growth, the coherence between hydrological and meteorological drought reduced, the lagging effect of hydrological drought attenuated, and the hydrological one tended to be more synchronized with meteorological one over the long period; and (3) the larger the watershed scale, the later (earlier) occurrence of the hydrological drought onset (termination), while the longer drought duration and larger magnitude for triggering hydrological drought.