Tag: Climate Change Downscaling Techniques

Abstract

Several studies have shown that climate change may enhance the severity of droughts over the Western Cape (South Africa) in the future, but there is a dearth of information on how to reduce the impacts of climate change on water yields. This study investigates the extent to which land-use changes can reduce the projected impacts of climate change on hydrological droughts in the Western Cape catchments. For the study, the Soil Water Assessment Tool (SWAT +) model was calibrated and evaluated over several river catchments, and the climate simulation dataset from the COordinated Regional Downscaling EXperiment (CORDEX) was bias-corrected. Using the bias-corrected climate data as a forcing, the SWAT + was used to project the impacts of future climate change on water yield in the catchments and to quantify the sensitivity of the projection to four feasible land-use change scenarios in the catchments. The land-use scenarios are the spread of forest (FOMI), the restoration of shrubland (SHRB), the expansion of cropland (CRDY), and the restoration of grassland (GRSL).

The model evaluation shows a good agreement between the simulated and observed monthly streamflows at four stations, and the bias correction of the CORDEX dataset improved the hydrological simulations. The climate change projection features an increase in temperature and potential evaporation, but a decrease in precipitation and all the hydrological variables. The drying occurs across the Western Cape, with the magnitude increasing with higher global warming levels (GWLs). The land-use changes alter these climate change impacts through changes in the hydrological water balance. FOMI increases streamflow and decreases runoff, while SHRB decreases streamflow and runoff. The influence of CRDY and GRSL are more complex. However, all the impacts of land-use changes are negligible compared to the impacts of climate change. Hence, land-use changes in the Western Cape may not be the most efficient strategies for mitigating the impacts of climate change on hydrological droughts over the region. The results of the study have application towards improving water security in the Western Cape river catchments.

Abstract

Recent advancements have seen a pervasive application of machine learning methodologies in assessing the susceptibility of geological hazards. A pivotal element influencing the accuracy of model predictions resides in the prudent selection of model parameters within machine learning frameworks. The objective of this study is to develop a robust landslide susceptibility assessment model by refining the support vector machine (SVM) model through the employment of the Bayesian algorithm for hyperparameter optimization. The southern part of the Qinghai-Tibet Plateau, focusing on major highways, is selected as the study area. Nine influencing factors, namely the elevation, slope, aspect, profile curvature, lithology, topographic wetness index, normalized difference vegetation index, distance to faults, and distance to rivers, are selected as the conditioning variables instrumental in evaluating the likelihood of collapse occurrences. Secondly, data from field surveys involving 351 landslides and randomly generated non-landslide data are utilized in a balanced 1:1 ratio to construct the training and testing datasets. Next, the cross-validation loss rate of the SVM model is selected as the objective function, and the Bayesian algorithm is used to optimize the BoxConstraint and KernelScale parameters of the SVM model, resulting in a Bayesian optimization-based SVM model. The results show that, within a five-fold cross-validation framework, the model yields 99.15% and 96.32% accuracy for the training and testing datasets, respectively. Concurrently, the area under the receiver operating characteristic curve values are recorded at 99.76% and 98.67% for the respective datasets, highlighting a notable level of predictive proficiency. Furthermore, factor importance ranking reveals lithology and elevation as the most influential, with partial dependence plots identifying high susceptibility areas between elevations of 2916 and 3954 m under soft lithology conditions. A collapse susceptibility map encompassing the entire study area is encompassing, categorizing the study area into extremely high (7.79%), high (13.38%), moderate (29.99%), and low (48.84%) susceptibility zones.

Abstract

Mapping dynamically distributed livestock in the vast steppe area based on statistical data collected by administrative units is very difficult as it is limited by the quality of statistical data and local geographical environment factors. While, spatial mapping of livestock gridded data is critical and necessary for animal husbandry management, which can be easily integrated and analyzed with other natural environment data. Facing this challenge, this study introduces a spatialization method using random forest (RF) in the Selenge River Basin, which is the main animal husbandry region in Mongolia. A spatialized model was constructed based on the RF to obtain high-resolution gridded distribution data of total livestock, sheep & goats, cattle, and horses. The contribution of factors influencing the spatial distribution of livestock was quantitatively analyzed. The predicted results showed that (1) it has high livestock densities in the southwestern regions and low in the northern regions of the Selenge River Basin; (2) the sheep & goats density was mainly concentrated in 0–125 sheep/km², and the high-density area was mainly distributed in Khuvsgul, Arkhangai, Bulgan and part soums of Orkhon; (3) horses and cattle density were concentrated in 0–25 head/km², mainly distributed in the southwest and central parts of the basin, with few high-density areas. This indicates that the RF simulation results effectively depict the characteristics of Selenge River Basin. Further study supported by Geodetector showed human activity was the main driver of livestock distribution in the basin. This study is expected to provide fundamental support for the precise regulation of animal husbandry in the Mongolian Plateau or other large steppe regions worldwide.

Abstract

Efficient drought monitoring in the Urmia Lake basin (ULB) is imperative to protect its ecosystem, agriculture, and the livelihoods of local communities relying on its water resources, considering the lake's susceptibility to changes in water availability. This study presents a novel approach to address the pressing issue of precise drought monitoring in regions with limited and unevenly distributed weather stations. By utilizing Tropical Rainfall Measuring Mission (TRMM) satellite-derived data and a Geographically Weighted Regression (GWR) model, we have significantly refined the spatial resolution of TRMM-3B43 precipitation (5 km) compared with the original TRMM data (about 25 km). This innovative methodology, which uses the globally available MODIS data of Normalized Difference Vegetation Index (NDVI) and day-night difference in land surface temperature (LST_dn) as independent variables, achieves spatial downscaling, enhancing the spatial resolution to 5 km for the period 2001–2019. The downscaled precipitation data from three models (M1: GWR_NDVI, M2: GWR_LST_dn, and M3: GWR_NDVI-LST_dn) were applied for drought assessment based on the Standardized Precipitation Index (SPI) and modified Rainfall Anomaly Index (mRAI). The findings demonstrate that the M2 model is more accurate than the other models for spatial downscaling of TRMM 3B43 data, with reductions in RMSE and MAE values by 6.04 and 4.16 mm, respectively. Additionally, this downscaling model significantly improves KGE values from 0.56 to 0.92 while achieving the lowest percentage bias. Monthly downscaled TRMM data based on LST across 12 synoptic stations in the basin reveals enhanced accuracy post-downscaling. Spatial analysis of average monthly precipitation maps illustrates a descending rainfall trend from June to September, followed by an ascending trend from October, with peak rainfall in November and December, notably in the western and southwestern regions. This analysis of precipitation trends offers valuable insights into the spatial distribution of rainfall within the basin, revealing variations across regions crucial for effective water resource management. The concentration of higher precipitation levels in the western and southwestern sectors underscores the significance of targeted water conservation and storage efforts. Drought severity analysis unveiled persistent and escalating drought conditions, causing significant impacts across the basin. A comparative assessment of severity classes using SPI and mRAI indices at 12 synoptic stations demonstrated strong agreement. Drought occurrences were a near-annual affair, attaining severe levels during specific years, notably 2008, 2011, and 2019. Spatial analysis revealed widespread drought events affecting nearly half of the basin area, with a noticeable worsening in severity during critical years. This highlights the practical need for adopting drought mitigation strategies and water resources management.

Abstract

The global population is aging at the same time as heat exposures are increasing due to climate change. Age structure, and its biological and socio-economic drivers, determine populations’ vulnerability to high temperatures. Here we combine age-stratified demographic projections with downscaled temperature projections to mid-century and find that chronic exposure to heat doubles across all warming scenarios. Moreover, >23% of the global population aged 69+ will inhabit climates whose 95th percentile of daily maximum temperature exceeds the critical threshold of 37.5 °C, compared with 14% today, exposing an additional 177–246 million older adults to dangerous acute heat. Effects are most severe in Asia and Africa, which also have the lowest adaptive capacity. Our results facilitate regional heat risk assessments and inform public health decision-making.

Abstract

Climate models are fundamental tools to estimates the reliable future climate change and its effects on the water resources and agriculture in basins. However, all climate models are not equally performed for all areas. Therefore, determining the most appropriate climate models for a specific study area is essential. The focus of this study was to evaluate the performance of the regional climate models with regard to simulating precipitation, and temperatures at Katar watershed. This study examines the performance of fourteen CORDEX-AFRICA-220 Regional Climate Models (RCMs) for the period of 1984–2005 using statistical metrics such as Pearson correlation coefficient (R), mean absolute error (MAE), root mean squared error (RMSE), and bias. The findings indicated that GERICS-MPI was better performed in representing Areta, and Bokoji station, GERICS-IPSL was better representing in Assela, Ketergenet, and Sagure station, CCCma-CanESM2-AFR22, and RCA4-ICHEC performed relatively better in representing the mean annual observed rainfall at the Kulumsa, and Ogolcho station respectively. However, RCA4-CSIRO performed weakly in estimation of annual rainfall at all stations. RCM model such as GERICS-MPI was relatively better than the others in replicating the annual pattern of the maximum temperature at Areta, Bokoji, and Ketergenet stations. Similarly, GERICS-IPSL were relatively better in replicating the annual maximum temperature at Assela, and Sagure stations, CCCma-CanESM2-AFR22 at Kulumsa station, and RCA4-ICHEC at Ogolcho station performed well in capturing the observed and simulated annual maximum temperature. Better performance was observed on minimum temperature at CCCma-CanESM2-AFR22 at Areta, Assela, and Ketergenet stations, GERICS-MOHE-AFR-22 at Bokoji station, GERICS-MPI at Kulumsa, and Ogolcho stations, RAC4-NOAA-2G at Sagure stations. However, weak performance was observed RCA4-CSIRO at all stations. RCM models of GERICS-MPI, and CCLM4-NCC-AFR-22 performed better than the other RCM models for correction of annual rainfall in Katar watershed. However, poor performance was observed at RCA4-ICHEC model on Katar watershed. The GERICS-MPI model performed well. However, poor performance was observed at RCA4-ICHEC on maximum temperature, and GERICS-NOAA-2M on minimum temperature in Katar watershed.

Abstract

Meteorological processes over islands with complex orography could be better simulated by Convection Permitting Regional Climate Models (CP-RCMs) thanks to an improved representation of the orography, land–sea contrasts, the combination of coastal and orographic effects, and explicit deep convection. This paper evaluates the ability of the CP-RCM CNRM-AROME (2.5-km horizontal resolution) to simulate relevant meteorological characteristics of the Mediterranean island of Corsica for the 2000–2018 period. These hindcast simulations are compared to their driving Regional Climate Model (RCM) CNRM-ALADIN (12.5-km horizontal resolution and parameterised convection), weather stations for precipitation and wind and gridded precipitation datasets. The main benefits are found in the representation of (i) precipitation extremes resulting mainly from mesoscale convective systems affected by steep mountains during autumn and (ii) the formation of convection through thermally induced diurnal circulations and their interaction with the orography during summer. Simulations of hourly precipitation extremes, the diurnal cycle of precipitation, the distribution of precipitation intensities, the duration of precipitation events, and sea breezes are all improved in the 2.5-km simulations with respect to the RCM, confirming an added value. However, existing differences between model simulations and observations are difficult to explain as the main biases are related to the availability and quality of observations, particularly at high elevations. Overall, better results from the 2.5-km resolution, increase our confidence in CP-RCMs to investigate future climate projections for Corsica and islands with complex terrain.

Abstract

In response to the growing demand for high-resolution rainfall data to support disaster prevention in Taiwan, this study presents an innovative approach for downscaling precipitation data. We employed a hierarchical architecture of Multi-Scale Residual Networks (MSRN) to downscale rainfall from a coarse 0.25-degree resolution to a fine 0.0125-degree resolution, representing a substantial challenge due to a resolution increase of over 20 times. Our results demonstrate that the hierarchical MSRN outperforms both the one-step MSRN and linear interpolation methods when reconstructing high-resolution daily rainfall. It surpasses the linear interpolation method by 15.1 and 9.1% in terms of mean absolute error and root mean square error, respectively. Furthermore, the hierarchical MSRN excels in accurately reproducing high-resolution rainfall for various rainfall thresholds, displaying minimal biases. The threat score (TS) highlights the hierarchical MSRN's capability to replicate extreme rainfall events, achieving TS scores exceeding 0.54 and 0.46 at rainfall thresholds of 350 and 500 mm per day, outperforming alternative methods. This method is also applied to an operational global model, the ECMWF’s daily rainfall forecasts over Taiwan. The evaluation results indicate that our approach is effective at improving rainfall forecasts for thresholds greater than 100 mm per day, with more significant improvement for the 1- to 3-day lead forecast. This approach also offers a realistic visual representation of fine-grained rainfall distribution, showing promise for making significant contributions to disaster preparedness and weather forecasting in Taiwan.

Abstract

NCT Delhi, the heart of India, remains vulnerable to urban flooding from July to October, when southwest monsoon is active over its area of 1483 km². To address the paucity of a comprehensive susceptibility map, this study employs urban flood modeling to quantify the spatial sensitivity of NCT Delhi to water logging. Fifteen flood related variables, including elevation, slope, slope aspect, profile curvature, plan curvature, Topographic Wetness Index (TWI), Stream Power Index (SPI), Topographic Roughness Index (TRI), geology, soil type, land use/land cover (LU/LC), Modified Fournier Index (MFI), water level depth, distance from storm drain, and distance from the Yamuna were analyzed. The study applies the data-driven form of ensemble fuzzy weights of evidence-support vector machine (Fuzzy WofE-SVM). The first-level flood susceptibility map was generated using Fuzzy WofE technique. Using this map and secondary data on water logging locations identified by the Delhi Traffic Police, flood inventory datasets were created. To execute the ensemble methodology, SVM was sequentially integrated with the Fuzzy WofE technique. Urban flood susceptibility zonation maps of NCT Delhi were created using four SVM kernel functions: linear (LN), polynomial (PL), radial basis function (RBF), and sigmoid (SIG). The kernel parameters i.e., regularization parameter (C), kernel width (γ) and degree (d) were optimized using python-mediated k-fold cross validation method. Area under the receiver operating characteristic curve (AUROC) analysis validated that all the SVM kernels yield excellent success and prediction rates. However, in terms of success rate, RBF (AUROC = 0.976) outperforms PL (AUROC = 0.968), LN (AUROC = 0.966), and SIG (AUROC = 0.956). With an AUROC of 0.966, RBF again outperforms SIG (AUROC = 0.964), LN (AUROC = 0.955), and PL (AUROC = 0.952) in terms of predictive performance. The novel Coupled Risk (CR) Index has been developed and presented in this study to increase the applicability of flood modeling by translating macro-hazard perspectives into a polished vulnerability scenario of administrative convenience. Spatial analysis of hazard intensity and endangered population using this novel tool would benefit take specific actions for disaster management by prioritizing wards for mitigation strategy planning and implementation.

Abstract

Climate change and rising temperatures represent a grave threat to modern civilization, with many of the key infrastructures across the world under risk. The energy sector is highly susceptible to anthropogenic warming, and many of its subsectors would be impacted by it. We use climate projections from climate models and use them to estimate the impact on demand, transmission, and generation in the country. An ANN-based approach utilizing historical demand patterns is used to estimate the impact on consumer demand, and it showed that most regions in the country would experience a drastic increase in demand. A thermal model of transmission lines showed that some transmission lines might lose 23.34% of their capacity. Data reveals that renewable energy sources boost energy efficiency. Hence, future national policies should include more of them. Thermal and PV power in the country have been compared to their resilience to rising temperatures. Without the implementation of more efficient technologies, demand-side management programs, or the upgrading of transmission infrastructure, CC impacts may overwhelm Pakistan's already weak grid system. Our applied models show the Highest forecasting efficiency to be 92.42%, 92.02% and 91.98% for PESCO, LESCO and IESCO.