Tag: Climate Change Downscaling Techniques

Abstract

Surface soil moisture plays a crucial role in various fields such as climate change, agronomy, water resources, and many other scientific and engineering domains. Accurately measuring soil moisture at both regional and global scales, with high spatial and temporal resolution, is essential for predicting and managing floods, droughts, and agricultural productivity to ensure food security. The launch of Sentinel operational satellites has significantly advanced remote sensing observations, enabling scientists to estimate soil moisture more accurately at improved spatial and temporal resolutions. This study aims to assess the potential of utilizing Sentinel-1A satellite images for soil moisture estimation in a semi-arid region using the Modified Dubois Model (MDM) semi-empirical model with Topp’s model. The soil moisture estimated is validated by comparing it with field measurements, which helps in understanding the spatial variability of soil moisture across various land use classes. Results concluded that the Sentinel 1 derived soil moisture on 3rd and 15th January 2022 in comparison with the soil moisture measured using soil moisture probe (R² = 0.68 and 0.63) and laboratory measurement (R² = 0.72 and 0.72) are found to be well correlated and can be adapted for monitoring drought and managing water resources. The study offers a robust accuracy assessment of Sentinel 1 derived soil moisture using soil moisture probe and laboratory analysis and suggests that the framework has the potential for operational monitoring of drought conditions and water resource management in semi-arid regions at a higher spatial and temporal resolution.

Abstract

Seasonal rainfall plays a vital role in both environmental dynamics and decision-making for rainfed agriculture in Ethiopia, a country often impacted by extreme climate events such as drought and flooding. Predicting the onset of the rainy season and providing localized rainfall forecasts for Ethiopia is challenging due to the changing spatiotemporal patterns and the country’s rugged topography. The Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), ERA5-Land total precipitation and temperature data are used from 1981–2022 to predict spatial rainfall by applying an artificial neural network (ANN). The recurrent neural network (RNN) is a nonlinear autoregressive network with exogenous input (NARX), which includes feed-forward connections and multiple network layers, employing the Levenberg Marquait algorithm. This method is applied to downscale data from the European Centre for Medium-range Weather Forecasts fifth-generation seasonal forecast system (ECMWF-SEAS5) and the Euro-Mediterranean Centre for Climate Change (CMCC) to the specific locations of rainfall stations in Ethiopia for the period 1980–2020. Across the stations, the results of NARX exhibit strong associations and reduced errors. The statistical results indicate that, except for the southwestern Ethiopian highlands, the downscaled monthly precipitation data exhibits high skill scores compared to the station records, demonstrating the effectiveness of the NARX approach for predicting local seasonal rainfall in Ethiopia’s complex terrain. In addition to this spatial ANN of the summer season precipitation, temperature, as well as the combination of these two variables, show promising results.

Abstract

Understanding the characteristics of extreme rainfall is vital for planning effective adaptation and mitigation measures. Thus, this study aims to evaluate the performances of 16 general circulation models (GCMs) of the coupled model intercomparison project (8 CMIP5 and 8 CMIP6) in reproducing observed extreme rainfall indices (ERFIs) and monthly rainfall in the Upper Blue Nile (UBN) basin of Ethiopia (1981–2005). The observed ERFIs were computed based on rainfall estimates of the random forest merging (RF-MERGE) algorithm, which combines ground-based rainfall with three gridded rainfall products (GRFPs). The GCMs were evaluated using statistical performance measures such as the Pearson correlation coefficient (R), root mean square error (RMSE), and percent bias (PBIAS). Using the Technique for Order Preferences by Similarity to an Ideal Solution (TOPSIS), the GCMs were ranked from least skilled to most skilled. Accordingly, MIROC5 of the CMIP5 and MPI-ESM1-2-LR of the CMIP6 models were found to be the most suitable models. Furthermore, the top-ranked models were selected and bias corrected using quantile mapping (QM), and their ensembles (CMIP5-ensemble and CMIP6-ensemble) were used for projecting future extremes under representative concentration pathways (RCP4.5 and RCP8.5) and shared socioeconomic pathways (SSP2-4.5 and SSP5-8.5), respectively, for the periods 2031–2055 and 2056–2080. Most of the ERFIs exhibited high variability and inconsistent trends for both the observations and future periods. The findings of this study provide valuable insights the the impacts of climate change on ERFIs, and the developed framework can serves as a useful reference for future research.

Abstract

Global environmental changes, such as climate change and reclamation alterations, significantly influence hydrological processes, leading to hydrologic nonstationarity and challenges in managing water availability and distribution. This study introduces a conceptual underpinning for the rational development and sustainability of groundwater resources. As one of the areas intended for the development projects within the Egyptian national plan for the reclamation of one and a half million acres; hundreds of pumping wells were constructed in the Moghra area to fulfill the reclamation demand. This study investigates the long-term impacts of exploiting the drilled pumping wells under climate change. The approach is to monitor the groundwater levels and the salinity values in the Moghra aquifer with various operational strategies and present proposed sustainable development scenarios. The impact of global warming and climate change is estimated for a prediction period of 30 years by using satellite data, time series geographical analysis, and statistical modeling. Using MODFLOW and Solute Transport (MT3DMS) modules of Visual MODFLOW USGS 2005 software, a three-dimensional (3D) finite-difference model is created to simulate groundwater flow and salinity distribution in the Moghra aquifer with the input of forecast downscaling (2020–2050) of main climatic parameters (PPT, ET, and Temp). The optimal adaptation-integrated scenario to cope with long-term groundwater withdrawal and climate change impacts is achieved when the Ministry of Irrigation and Water Resources (MWRI) recommends that the maximum drawdown shouldn’t be more significant than 1.0 m/ year. In this scenario, 1,500 pumping wells are distributed with an equal space of 500 m, a pumping rate of 1,200 m³/day and input the forecast of the most significant climatic parameters after 30 years. The output results of this scenario revealed a drawdown level of 42 m and a groundwater salinity value of 16,000 mg/l. Climate change has an evident impact on groundwater quantity and quality, particularly in the unconfined coastal aquifer, which is vulnerable to saltwater intrusion and pollution of drinking water resources. The relationship between climate change and the hydrologic cycle is crucial for predicting future water availability and addressing water-related issues.

Abstract

Projected changes in atmospheric blocking and associated extreme weather are marked by considerable uncertainties. While paleoclimate records could help reduce these uncertainties, their low temporal resolution makes extracting synoptic-scale signals challenging. Here, a deep learning model is developed to infer summertime blocking frequency from tree-ring-based gridded reconstructions of Northern Hemisphere surface temperature over the Last Millennium. The model, despite not directly incorporating paleoclimate proxies or their locations, is implicitly constrained by them. The reconstructions highlight the tropical Pacific’s strong influence on blocking variability at interannual-to-centennial time scales. A weakened tropical Pacific zonal temperature gradient during the Little Ice Age correlates with a hemispherically reduced -yet more variable interannually- blocking frequency and altered regional patterns. This deep learning approach offers a pathway for extracting paleoweather signals from paleoclimate records that enables improved understanding of blocking response to external forcing and constraining of model projections of blocking under climate change scenarios.

Abstract

Unmanned aerial vehicle (UAV) imaging methods have drawn a lot of interest lately from academics and industry professionals as an affordable option for agro-environmental uses. To improve the UAV capabilities of diverse applications, machine learning (ML) methods are specifically applied to UAV-based remote sensing data. Spatiotemporal properties were analysed and the city-level carbon emissions statistics were estimated. This research proposes novel technique in UAV-based climate temperature pattern analysis and carbon emission detection utilizing the deep learning (DL) model. Here the input is collected as UAV-based weather data which is processed for noise removal and smoothening. Processed data is extracted and classified utilizing Gaussian belief deep neural network and spatial convolutional Q-swarm colony metaheuristic optimization. A metaheuristic optimisation algorithm uses gradient information-free iterative evaluation of the objective function in order to identify a global optimum. Experimental analysis has been carried out in terms of detection accuracy, average precision, F-1 score, recall and AUC for various UAV-based weather dataset. The proposed technique attained mean average precision was 92%, recall was 94%, AUC was 87%, detection accuracy was 96% and F1-score was 93%. This work shows that remotely sensed data can be used to support more advanced evaluations that are more successful, particularly in areas with extensive selective logging and diverse forest conditions, and to help quantify carbon emissions from selective logging using conventional methodologies.

Abstract

Assessing and selecting climate models with lower uncertainty is necessary to predict future climate and hydrological risks at the watershed scale. In this study, we integrated stepwise cluster analysis (SCA) to propose a multi-model ensemble downscaling framework aimed at reducing the uncertainty of GCM-based precipitation projections in large-scale watersheds. The Pearl River Basin (PRB) in southern China was selected as the study area to validate the reliability of this framework. Spatially, we investigated the features of terrain-related spatial heterogeneity in precipitation simulation of different climate models using a stepwise cluster zoning approach. The spatial performance of most CMIP6 models was effective in capturing the annual mean precipitation from the source region to the downstream of the PRB. To further evaluate the model's skill in simulating precipitation patterns, we conducted a seasonal analysis for different periods throughout the year. However, the seasonal precipitation cycle exhibited a wet bias during cold seasons, and the most significant deviation of precipitation percentage intervals occurred during winter. The TSS ranking of CMIP6 models was used to select the top-performing models to construct an improved multi-model ensemble mean (MEM5), resulting in a more accurate precipitation simulation for PRB. Results showed consistent precipitation increases (p < 0.05) for all scenarios in the PRB, with the middle and lower reaches being the most sensitive to changes in precipitation. The improved MEM5 can serve as a valuable reference for accurately simulating hydrological regimes and extreme weather events in the PRB. The proposed multi-model ensemble downscaling framework, which incorporates SCA, offers a new approach for high-resolution and low-uncertainty climate simulations in other large-scale watersheds.

Abstract

Located in the climatically vulnerable Middle East, Iraq is projected to be highly susceptible to climate change impacts. Evaluating these effects on water resources requires an analysis of key hydrological factors, including rainfall and associated runoff. The Soil Conservation Service Curve Number (SCS-CN) method remains the foremost runoff modelling approach because it relies on the curve number (CN), derived from land use, soil type, and other parameters. In this chapter, the SCS-CN model is integrated with geographic information systems (GIS) and Long Aston Research Station Weather Generation (LARS-WG) to estimate rainfall-runoff under climate change projections for the Kirkuk governorate in Iraq. The SCS-CN model inputs of hydrological soil properties, rainfall data, and potential retention were obtained through GIS analysis. The results from 1992 to 2022 showed an average annual runoff of a maximum runoff of 290 mm in 1992 and a minimum of 184 mm in 2021. Under future climate change projections, the estimated annual average runoff is predicted to decline to 110 mm by 2055, with a maximum of 280 mm by 2025. These findings indicate substantial climate change impacts on future rainfall, the associated surface runoff, water availability, and water resources in the Kirkuk governorate. Integrating established hydrologic models, such as SCS-CN with GIS, can enhance climate change impact assessments for improved water resource planning.

Abstract

Recent development of artificial intelligence (AI) technology has resulted in the fruition of machine learning-based weather prediction (MLWP) systems. Five prominent global MLWP model, Pangu-Weather, FourCastNet v2 (FCN2), GraphCast, FuXi, and FengWu, emerged. This study conducts a homogeneous comparison of these models utilizing identical initial conditions from ERA5. The performance is evaluated in the Eastern Asia and Western Pacific from June to November 2023. The evaluation comprises Root Mean Square Error and Anomaly Correlation Coefficients within the designated region, typhoon track and intensity predictions, and a case study for Typhoon Haikui. Results indicate that FengWu emerges as the best-performing model, followed by FuXi and GraphCast, with FCN2 and Pangu-Weather ranking lower. A multi-model ensemble, constructed by averaging predictions from the five models, demonstrates superior performance, rivaling that of FengWu. For the 11 typhoons in 2023, FengWu demonstrates the most accurate track prediction; however, it also has the largest intensity errors.

Abstract

The stability of environmental conditions around the world has been threatened by climate change in the last few decades. Sea levels are rising more quickly than ever before due to a combination of factors including temperature increases, changes in precipitation patterns, and glacier melting. We describe a machine learning strategy to design coastal sea level fluctuation as well as related uncertainty across a range of timescales by utilising important ocean temperature estimations as proxies for regional thermosteric sea level component. The aim of this research is to propose novel method in climate change with water level rise analysis using biosensor with machine learning techniques in diverse biological ecosystems. Here, the input is collected as biosensor-based climate analysis with water level rise analysis dataset and processed for noise removal, normalisation, and smoothening. Then, climate data analysis is carried out utilising fuzzy adversarial encoder (FAE) model and the water level rise analysis is carried out using recurrent transfer AlexNet neural network (RTAlexNetNN). The classified output shows climate change analysis with water level rise modelling. The experimental analysis has been carried out for various climate data and water rise data in terms of random accuracy, specificity, MSE, F-measure, and normalised cross-correlation. The proposed technique obtained 98% random accuracy, 96% normalised cross-correlation, 94% specificity, 58% MSE, and 92% F-measure.