Using the MODIS Sensor for Snow Cover Modeling and the Assessment of Drought Effects on Snow Cover in a Mountainous Area

Exploring the comprehensive link between climatic factors and vegetation productivity in China

Understanding the influence of climatic factors on vegetation dynamics and cumulative effects is critical for global sustainable development. However, the response of vegetation to climate and the underlying mechanisms in different climatic zones remains unclear. In this study, we analyzed the response of vegetation gross primary productivity (GPP) to climatic factors and the cumulative effects across various vegetation types and climatic zones, utilizing data on precipitation (Pr), temperature (Ta), and the standardized precipitation evapotranspiration index (SPEI). The results showed that: (1) GPP showed significant differences among the seven climatic zones, with the highest value observed in zone VII, reaching 1860.07 gC·m- 2, and the lowest in zone I, at 126.03 gC·m- 2. (2) GPP was significantly and positively correlated with temperature in climatic zones I, IV, V, and VI and with precipitation in climatic zones I, II, and IV. Additionally, a significant positive correlated was found between SPEI and GPP in climatic zones I, II, and IV. (3) Drought exerted a cumulative effect on GPP in 45.10% of the regions within China, with an average cumulative duration of 5 months. These effects persisted for 6-8 months in zones I, II, and VII, and for 2-4 months in zones III, IV and VI. Among different vegetation types, forests experienced longest cumulative effect time of 6 months, followed by grasslands (5 months), croplands (4 months), and shrublands (4 months). The cumulative time scale decreased with increasing annual SPEI. The varying responses and accumulation of GPP to drought among different vegetation types in various climatic zones underscore the complexity of vegetation-climate interactions the response and accumulation of GPP to drought.

Spatiotemporal Variation of Snow Cover Frequency in the Qilian Mountains (Northwestern China) during 2000–2020 and Associated Circulation Mechanisms

Linking snow cover frequency (SCF) and atmospheric circulation is vital for comprehension of hemispheric-scale change mechanisms and for accurate forecasting. This study combined MODIS imagery with meteorological observations to investigate the variation of annual SCFs in the Qilian Mountains. Results indicated that more than 80% of annual SCF is distributed at high elevations and mostly on northern slopes, and that SCF is greater in the west than in the east. Abrupt change in the increase in annual SCF was not detected; however, significant (0.05 confidence level) variation with quasi-3-year and quasi-5-year periods indicated potential connection with monsoons. Topographically, SCF increased at high elevations and decreased in valleys. Moreover, SCF increased significantly with a rise in slope below 23° and then decreased between 23° and 45°, and it decreased with a change in aspect from 70° to 200° and then increased from 200° to 310°. Annual SCF variation in the Qilian Mountains is dominated by precipitation rather than by temperature. In the years with high SCFs, southeasterly winds associated with an anticyclone over southeastern China and southwesterly winds associated with the cyclone over the Iranian Plateau brought warm moisture across northwestern China, favoring snowfall in the Qilian Mountains. Meanwhile, cold moisture outbreaks from the Arctic into the mid-latitudes are conducive to maintaining snow cover. However, in the years with low SCFs, the cold air might be difficultly transporting out of the Arctic region due to the strengthening polar vortex. Moreover, the water vapor was less than that of the mean state and divergence over the Qilian Mountains, which difficultly conduced snowfall over the Qilian Mountains.

Accelerated decline of snow cover in China from 1979 to 2018 observed from space

Snow cover is an important indicator of climate change. Variations of snow integrate the competing effect of increasing temperature and precipitation. In this study, based on Theil-Sen Median (TSM) and Mann-Kendall (M-K) methods, observational evidence from space was used to investigate the variation of snow parameters in China from 1979 to 2018, and some meaningful conclusions were found. (1) The downward trend of snow depth (SD) with a median of -0.02 cm/year was generally found in the high altitude mountains, and the upward trend of SD with a median of 0.01 cm/year occurs in the plains. A widespread and accelerated decrease of SD was observed in the latest period (2009-2018) in NC and NX. (2) The mean annual areas with snow cover days (SCDs) greater than 150 days (d) accounted for 17.8%, 24.73% and 38.14% in NC, NX and QTP. SCDs in NC and Northern QTP were widely reduced, but the longest snow season with more than eight months is still maintained on QTP. (3) The downward trend of snow storage (SS) was found in all three snow areas. The difference of snow phenology is reflected in the slowest accumulation and melting rate of SS on QTP; the largest peak value of SS and the shortest snow season in NC; the slow accumulation and rapid melting of snow in the NX, and the peak value is achieved at the latest. The trend of maximum temperature was judged as the most important factor affecting the change of SD, while longitude and latitude are closely related to the change of SCD.

Comparing linear and non-linear data-driven approaches in monthly river flow prediction, based on the models SARIMA, LSSVM, ANFIS, and GMDH

River flow variations directly affect the hydro-climatological, environmental, and ecological characteristics of a region. Therefore, an accurate prediction of river flow can critically be important for water managers and planners. The present study aims to compare different data-driven models in predicting monthly flow. Two river catchments located in the Guilan province in Iran, where rivers play an essential role in agricultural productions (mainly rice), are studied. The monthly river flow dataset was provided by Guilan Regional Water Authority during 1986-2015. The models are derived from two different numerical types of stochastic and machine learning (ML) models. The stochastic model is seasonal autoregressive integrated moving average (SARIMA), and the MLs are least square support vector machine (LSSVM), adaptive neuro-fuzzy inference system (ANFIS), and group method of data handling (GMDH). The inputs were selected by autocorrelation and partial autocorrelation functions (ACF and PACF) from the flow rates of the previous months. The data was divided into 75% of training and 25% of testing phases, and then the mentioned models were implemented. Predictions were evaluated by the criteria of root mean square error (RMSE), normalized RMSE (NRMSE), and Nash Sutcliff (NS) coefficient. According to the calculated values of different criteria during the test phase, RMSE = 1.138 cms, NRMSE = 0.109, and NS = 0.826, it can be concluded that the SARIMA model was superior to its ML competitors. Among the ML models, GMDH had the best performance (by RMSE = 1.290 cms, NRMSE = 0.124, and NS = 0.777) because it has more optimization parameters and sample space for network make-up. The models were also evaluated in hydrological drought conditions of both rivers. It was resulted that the rivers' flow can be well predicted in drought conditions by using these models, especially the SARIMA stochastic model. According to the NRMSE values (ranged between 0.1 and 0.2), the accuracy of predictions is evaluated in the appropriate range, and the present study shows promising results of the current approaches. Consequently, a comparison between the performance of linear stochastic models and complex black-box MLs, reveals that linear stochastic models are more suitable for the current region's monthly river flow prediction.

Evaluating global ecosystem water use efficiency response to drought based on multi-model analysis

Drought has serious consequences on terrestrial ecosystems, particularly for their carbon and water processes. As an important indicator to examine the balance of ecosystem water and carbon cycles, ecosystem water use efficiency (WUE) has been widely used to investigate ecosystem responses to drought. However, the response of WUE to drought and the role of different ecosystem processes in controlling the response of WUE to drought are not well studied. In this paper, we used four WUE datasets from different remote sensing-driven (RS-driven) models and three drought indices (Standardized Precipitation Evapotranspiration Index, soil moisture anomaly index and water storage anomaly-based drought index) to comprehensively investigate the response of WUE to drought and its dominant ecosystem processes during the period of 2001-2018. The results showed the WUE datasets from four different RS-driven models had discrepancies in WUE temporal trends, particularly in tropical and subtropical forest and semi-arid regions. The Spearman correlation analysis revealed that the positive correlations between WUE and drought accounted for more than half of global vegetated lands, while negative relationship mainly occurred in the high latitude regions. We further explored the dominant ecosystem processes (represented by GPP and ET) in controlling WUE response to drought, and found ET controlled WUE-drought relationship in the high latitude areas and semi-arid/sub-humid regions, while GPP dominated it in tropical forest regions. Additionally, the effects of GPP and ET on controlling WUE response to drought were examined to change with different drought indices, especially in the semi-arid regions. Our study suggests multi-model analysis tend to reduce uncertainties in analyzing WUE response to drought caused by a single WUE data. Moreover, our results highlight the different role of ecosystem processes in controlling WUE response to drought and provide new information for the underlying mechanism of drought impacts on ecosystem water and carbon cycles.

A Theoretical Approach for Forecasting Different Types of Drought Simultaneously, Using Entropy Theory and Machine-Learning Methods

Precipitation deficit can affect different natural resources such as water, soil, rivers and plants, and cause meteorological, hydrological and agricultural droughts. Multivariate drought indexes can theoretically show the severity and weakness of various drought types simultaneously. This study introduces an approach for forecasting joint deficit index (JDI) and multivariate standardized precipitation index (MSPI) by using machine–learning methods and entropy theory. JDI and MSPI were calculated for the 1–12 months’ time window (JDI1–12 and MSPI1–12), using monthly precipitation data. The methods implemented for forecasting are group method of data handling (GMDH), generalized regression neural network (GRNN), least squared support vector machine (LSSVM), adaptive neuro-fuzzy inference system (ANFIS) and ANFIS optimized with three heuristic optimization algorithms, differential evolution (DE), genetic algorithm (GA) and particle swarm optimization (PSO) as meta-innovative methods (ANFIS-DE, ANFIS-GA and ANFIS-PSO). Monthly precipitation, monthly temperature and previous amounts of the index’s values were used as inputs to the models. Data from 10 synoptic stations situated in the widest climatic zone of Iran (extra arid-cold climate) were employed. Optimal model inputs were selected by gamma test and entropy theory. The evaluation results, which were given using mean absolute error (MAE), root mean squared error (RMSE) and Willmott index (WI), show that the machine learning and meta-innovative models can present acceptable forecasts of general drought’s conditions. The algorithms DE, GA and PSO, could improve the ANFIS’s performance by 39.4%, 38.7% and 22.6%, respectively. Among all the applied models, the GMDH shows the best forecasting accuracy with MAE = 0.280, RMSE = 0.374 and WI = 0.955. In addition, the models could forecast MSPI better than JDI in the majority of cases (stations). Among the two methods used to select the optimal inputs, it is difficult to select one as a better input selector, but according to the results, more attention can be paid to entropy theory in drought studies.

Simulation of Titicaca Lake Water Level Fluctuations Using Hybrid Machine Learning Technique Integrated with Grey Wolf Optimizer Algorithm

Lakes have an important role in storing water for drinking, producing hydroelectric power, and environmental, agricultural, and industrial uses. In order to optimize the use of lakes, precise prediction of the lake water level (LWL) is a main issue in water resources management. Due to the existence of nonlinear relations, uncertainty, and characteristics of the time series variables, the exact prediction of the lake water level is difficult. In this study the hybrid support vector regression (SVR) and the grey wolf algorithm (GWO) are used to predict lake water level fluctuations. Also, three types of data preprocessing methods, namely Principal component analysis, Random forest, and Relief algorithm were used for finding the best input variables for prediction LWL by the SVR and SVR-GWO models. Before the LWL simulation on monthly time step using the hybrid model, an evolutionary approach based on different monthly lags was conducted for determining the best mask of the input variables. Results showed that based on the random forest method, the best scenario of the inputs was Xt−1, Xt−2, Xt−3, Xt−4 for the SVR-GWO model. Also, the performance of the SVR-GWO model indicated that it could simulate the LWL with acceptable accuracy (with RMSE = 0.08 m, MAE = 0.06 m, and R2 = 0.96).