Future changes in atmospheric moisture and wind field using numerical simulations: implications for moisture transport and wind energy
- Autores: José Carlos Fernández Alvarez
- Directores de la Tesis: Raquel Nieto Muñiz (dir. tes.), Luis Gimeno Presa (dir. tes.)
- Lectura: En la Universidade de Vigo ( España ) en 2024
- Idioma: inglés
- Tribunal Calificador de la Tesis: Ricardo Francisco García Herrera (presid.), Ana María Durán Quesada (secret.), Ricardo Trigo (voc.)
- Programa de doctorado: Programa de Doctorado en Agua, Sostenibilidad y Desarrollo por la Universidad de Vigo
- Materias:
- Enlaces
- Tesis en acceso abierto en: Investigo
- Resumen
Since the end of the 19th century, human actions have impacted the atmosphere, oceans, and land, leading to global warming. Consequently, the global water cycle has undergone substantial widespread changes since the 20th century. A warmer climate increases the transport of water vapour into climate systems, affecting the frequency and intensity of hydrometeorological extreme events, such as heavy rainfall and droughts, in many regions. In addition, climate change is expected to affect the wind speed at a height of 10 m (V10), with significant geophysical and social impacts, such as affecting the feasibility of wind energy production, wave climate, storm surges, and onshore and offshore industries. Furthermore, modifications in the wind field influence variables such as soil humidity, evaporation, and water availability to determine arid and semi-arid conditions.
Owing to climate variability, significant decreases in water availability are expected in several areas. Given the fragility of water availability, there is a growing need to understand the changes induced by global warming to ensure a sufficient supply of water resources. It is important to determine the dependence between the moisture transport given by the moisture source-sink relationship and climate change, and how this influences continental precipitation for specific regions. In addition, changes in global atmospheric circulation may affect the behaviour of regional wind patterns and compromise wind energy production capacity. Future changes in the spatial and temporal distributions of V10 imply changes in the type of energy associated with this variable. Therefore, an analysis of future projections would improve the production capacity and energy efficiency with better planning and balancing of ideal regions for installation.
Currently, notable progress has been made in studies focusing on determining where precipitation originates and the moisture source-sink relationship, establishing climatology for most regions of the world. Many studies have provided detailed information on the role of synoptic-scale systems, particularly in Atmospheric Rivers (ARs). These meteorological systems are considered to be one of the main moisture transport mechanisms, along with low-level jet systems, tropical and extratropical cyclones, and monsoons. Specifically, most of them have focused on the intensity and frequency of ARs and, to a lesser extent, on the determination of the sources that cause the humidity they transport or their anomalous moisture uptake. However, few studies have analysed the moisture source-sink relationship in a future climate over long periods or for specific systems such as ARs.
The Lagrangian approach is a widely used methodology for studying the moisture source-sink relationship. Previous studies based on this approach have not aimed to quantify how climate change will influence the location and importance of moisture source regions, and how this will determine the transport of moisture to sinks (continental areas), and even fewer have used high-resolution regional models. Moreover, regarding future changes in wind fields and related variables, numerous studies have been conducted in the Atlantic Ocean region using the Global Climate Models (GCMs) of the Coupled Model Intercomparison Project Phase 5 (CMIP5) and Phase 6 (CMIP6). However, the low resolution of the Global Climate Models used does not provide detailed information on climate change and its impacts in certain areas (regional scale), particularly where the weather and climate are not homogeneous.
Therefore, this thesis focuses on moisture transport and V10 by performing regional simulations at a higher spatial resolution, considering the North Atlantic as the study region. This spatial resolution increase will allow better reproduction of physical processes at higher spatial resolutions up to scales that allow the representation of convection (a few kilometres) essential for precipitation modelling. Furthermore, the results of this thesis are intended to provide more detail on the influence of climate change and its impact regionally and will allow decision-makers to take action on a given future climate signal or change. Specifically, a dynamic downscaling methodology which uses the advantages of the Eulerian mesoscale Weather Research and Forecasting (WRF) with an ARW dynamic kernel (WRF-ARW) and the Lagrangian dispersion model FLEXPART-WRF (regional model adapted from the FLEXible PARTicle dispersion model (FLEXPART) for WRF) was considered. Therefore, this study aimed to determine future changes in atmospheric moisture and wind fields using numerical simulations and analyse the implications for moisture transport and wind energy in the North Atlantic region.
To address the proposed overall objective, a dynamic downscaling methodology using reanalysis data was previously evaluated. This evaluation focuses on the representation of the moisture source-sink pattern using reanalysis data for several configurations of the dispersion models FLEXPARTv10.3 and FLEXPART-WRF for a specific period and over the North Atlantic Ocean region. Importantly, original and free software was developed to process the outputs of these models and obtain the necessary fields for the different studies. Next, the dynamically reduced dataset and the results of the FLEXPART-WRF dispersion model were used to determine future changes in the anomalous moisture uptake of the ARs reaching the Iberian Peninsula and in the strength and location of their moisture sources. Finally, future changes in V10 in the North Atlantic Ocean and how the modifications of the wind pattern influence offshore wind energy production were studied for three important subregions on each side of the basin: the Atlantic coast of the Iberian Peninsula, the United States East Coast, and the Caribbean Sea region.
The data used in this thesis come from different sources: reanalysis data from the European Centre for Medium-Range Weather Forecasts (ERA-Interim (ERA-I) and ERA5), climate outputs from the Community Earth System Model Version 2 (CESM2), and an ensemble of several climate models. These data were used as the initial and boundary conditions for the WRF-ARW model. The WRF-ARW outputs were used as inputs to the FLEXPART-WRF model. The Shared Socioeconomic Paths (SSPs) used in this study are the SSP2-4.5, SSP3-7.0, and SSP5-8.5 (representing an increase of 4.5 W/m2 , 7.0 W/m2, 8.5 W/m2 , respectively). These scenarios present differences among themselves due to the level of radiative forcing projected at the end of the century, but at the same time, they show a wide spectrum of possibilities from a more positive scenario to the most extreme. A set of 30-year periods was used in the analysis of moisture transport: 2036 - 2065, mid-century (MC), 2071-2100, end-century (EC), and 1985-2014, historical reference period (HIST), using only the SSP5-8.5 scenario. For the analysis of the wind field and wind power density (WPD), the three SSPs were used but 5-year periods were considered: 2049-2053, mid-century (MC 5Y); 2096-2100, end-century (EC 5Y); and 2010-2014, historical period (HIST 5Y). Finally, the changes projected considering the differences in scenarios were determined as the difference in the expected pattern at MC and EC minus the HIST pattern.
A Lagrangian methodology that follows changes in specific humidity every 6 h was applied, considering the trajectories of the particles in the atmosphere. The residence time of water vapour in the atmosphere considered was 10 days to track the particles forward or backward in time. In the study of moisture transport, the North Atlantic and the Mediterranean Sea (NATL and, MED, respectively) were considered as moisture sources, even the Iberian Peninsula itself, to study the precedes of recycling, and the surrounding continental regions as their sinks. Moreover, to study the moisture associated with ARs, the anomalous moisture uptake over the ocean, specifically for landfalling ARs arriving on the Iberian Peninsula, was determined. Finally, to calculate WPD, the wind speed was extrapolated to a height of 120 m in a neutral atmosphere. They were extrapolated following a logarithmic wind profile, and the Atlantic coast of the Iberian Peninsula, the United States East Coast, and the Caribbean Sea region were considered the study regions.
A software is required to process the outputs of particle dispersion models. The TRansport Of water Vapor (TROVA) tool was implemented to facilitate the application of the different Lagrangian methodologies to the outputs of the FLEXPART and FLEXPART-WRF dispersion models. In addition, TROVA allows the consideration of several mesh resolutions for the input and output data and masks used for moisture transport studies. Therefore, it can be used for the outputs of these dispersion models forced by both ERA-I and ERA5 and allows the study of projected changes in the strength and location of moisture sources and sinks using climate model outputs from WRF simulations. Furthermore, TROVA has good computational efficiency owing to its development using parallel programming, which allows its use in supercomputers. Finally, it is a free software implemented in Fortran and Python and is available to the scientific community.
Once the software was developed, the evaluation of the moisture sources and sink patterns for the ERA5 data (at 1 and 0.5 of spatial resolution) was carried out using the FLEXPART model and the dynamic downscaling methodology explained above. It was obtained that the three configurations presented a correlation in the range of 0.4 to 0.6 using ERA5 for the representation of the moisture pattern that provides humidity on the Iberian Peninsula, compared to that of the control experiment, higher for FLEXPART outputs using ERA5 at 0.5 and lower for the configuration using FLEXPART-WRF and WRF-ARW outputs (FLEX-WRF). A few noticeable differences were observed for the configurations regarding the error, with the greater error observed in summer and the lowest in winter and autumn. Finally, FLEX-WRF provided a better representation of the moisture source pattern, with a predominance of mountainous orography during spring and summer. Notably, precipitation recycling processes are predominant during these seasons.
In the analysis of the Mediterranean Sea source moisture sinks, the three configurations showed low dispersion and high correlation values compared with the control experiment. A lower mean absolute error was obtained with respect to the values given for the IP. The mean square error increased in summer and decreased in autumn. Additionally, FLEX-WRF best represented the moisture sink pattern in regions of complex orography, with the best performance of all configurations. However, the contribution pattern associated with the Mediterranean Sea tended to be more constricted both in latitude and longitude, showing minor errors compared with the North Atlantic target region. For North Atlantic Ocean, the magnitudes of the absolute error and root mean square error showed values of up to 2.5 and 4.5 mm/day respectively. Furthermore, the correlations between the three configurations and the control experiment ranged from 0.4 and 0.7. The FLEX-WRF model showed little ability to determine the sink patterns associated with this source.
Once all the configurations were evaluated, future climate studies were conducted considering the same moisture sources and their associated sinks. Specifically, in the two future periods analysed, a general increase in humidity was projected to reach the surrounding continental areas of the North Atlantic Ocean and Mediterranean Sea. In the MC, a slight annual increase was observed in contributions from the North Atlantic Ocean basin and the western Mediterranean Sea. At end-century, a pattern was observed with values greater than the expected changes at MC. A significant increase in precipitation recycling processes is expected over the Iberian Peninsula during the winter months and varies annually from 2 % to 8 %. The results obtained regarding the increase in moisture contributions for the NATL and MED over the IP are compatible with the amplification of the Clausius Clapeyron relationship. On the other hand, greater contributions from the Mediterranean Sea to continental precipitation were projected over the areas south of the Mediterranean Sea during both analysed periods. This contribution is expected to increase, being lower over the Iberian Peninsula mostly in summer, with maximum values at the end of the century. It is highlighted that the contribution in Eastern Europe will decrease, and for Western Europe, it is only appreciated in summer and spring. The NATL source will increase its contribution over the east coast of North America (mainly in winter and autumn), with notable values in the British Isles (mainly in winter), but it is expected to decrease latitudinally in the northern areas of western Europe, the Iberian Peninsula, and the west coast of Africa.
