Prediction of Severe Drought Area Based on Random Forest: Using Satellite Image and Topography Data

Abstract

The uncertainty of drought forecasting based on past meteorological data is increasing because of climate change. However, agricultural droughts, associated with food resources and determined by soil moisture, must be predicted several months ahead for timely resource allocation. Accordingly, we designed a severe drought area prediction (SDAP) model for short-term drought without meteorological data. The predictions of our proposed SDAP model indicate a forecast of serious drought areas assuming non-rainfall, not a probability prediction of drought occurrence. Furthermore, this prediction provides more practical information to help with rapid water allocation during a real drought. The model structure using remote sensing data consists of two parts. First, the drought function f(x) from the training area by random forest (RF) learned the changes in the pattern of soil moisture index (SMI) from the past drought and the training performance was found to be root mean square error (RMSE) = 0.052, mean absolute error (MAE) = 0.039, R 2 = 0.91. Second, derived f(x) predicted the SMI of the study area, which is 20 times larger than the training area, of the same season of another year as RMSE = 0.382, MAE = 0.375, R 2 = 0.58. We also obtained the variable importance stemming from RF and discussed its meaning along with the advantages and limitations of the model, training areas selection, and prediction coverage.Water 2019, 11, 705 2 of 15 resources. Droughts are generally classified into four categories, namely, meteorological, hydrological, agricultural, and socio-economic droughts [9].In recent years, the prevalence of machine learning methodologies, and frequent droughts and floods around the world, have increased the prediction of agricultural drought. However, the uncertainties of prediction caused by combination of meteorological factors [10][11][12][13], still remain a problem. According to the Intergovernmental Panel on Climate Change (IPCC) AR5 guidance note, the complex use of different models, complexity of models, and inclusion of additional processes in the analysis are the main reasons for the increase in uncertainties [14]. Thus, the complex models used for the integrated analysis of meteorological and agricultural drought have increased the uncertainty in the results [14]. However, agricultural drought prediction methods that do not include spatial information, and are only based on precipitation, such those put forth in [12,15], cannot predict the spatial distribution of agricultural drought. Another difficulty in predicting agricultural drought stems from predictions based on the meteorological pattern, such as patterns of past precipitation. Even well-designed agricultural drought models, based on meteorological factors, cannot make accurate predictions because of the shifts in existing patterns, due to climate change. According to a study of agricultural forecasting models from 2007 to 2017, (see Table 1 in [16]), precipitation was used as an input variable in almost all the models.Previous studies ...

Is Water Pricing Policy Adequate to Reduce Water Demand for Drought Mitigation in Korea?

“…This study has the framework of two linked models, as shown in Figure 2, which is a structure that simulates policy by SD model based on the result of SDAP model [13]. The SDAP model ( Figure 3) predicts the spatial distribution pattern of soil moisture after non-rainfall period using drought function trained by random forest (RF) algorithm [13]. The SD model (Figure 4) estimates the amount of water available to the provincial government by simulating the price increase policy for the drought-tolerant areas predicted in the previous process.…”

“…Agricultural drought was trained and predicted by the following concept [13]. The soil moisture after non-rainfall periods remains different depending on the condition of the present land surface [20].…”

“…The soil moisture after non-rainfall periods remains different depending on the condition of the present land surface [20]. In this regard, we classified the land surface factors that affect the soil moisture into four categories: vegetation, topographic, water, and thermal factors during the non-rainfall period [13]. Thermal factors reduce soil moisture, whereas vegetation retards the loss of soil moisture by slowing down the increase in land surface heat [21].…”

“…Table 1 shows the 15 features (variables) that correspond to four land surface factor. These 15 features are regressed on the soil moisture index (SMI) of three months later of no precipitation and it is the input variables for the RF regression [13]. Table 1.…”

Application of TRMM Precipitation Data to Evaluate Drought and Its Effects on Water Resources Instability

“…where R, σ P , and σ O are, respectively, correlation coefficient between TRMM and observational precipitation, standard deviation of TRMM precipitation data, and standard deviation of observational precipitation data. The NRMSE represents the mean square error dimensionless, and it is calculated by Equation (5) [21]:…”