Get Complete Project Material File(s) Now! »
Water resources in regional scale
At the regional scale, water resources are often considered within a river basin, within which the water budget can be closed (groundwater permitting). Humans directly change the dynamics of the water cycle through dams that are constructed for water storage or hydropower, and through water withdrawals for industrial, agricultural, or domestic purposes (Haddeland et al., 2014). Everything we do a ects the river basin, while in reverse, water resources in the river basin a ect humans by determining the food yield, the diversity of animal species, the speed of economic growth, and the number of people and livestock that can be sustainably supported. Among all the human activities, agriculture is the dominant component of human water use. Hence, many of the solutions to water-related food and environmental security problems come from the increasing of the e ciency of water use for the productivity of agriculture (Sophocleous, 2004b).
River basins in arid and semi-arid areas face a more signi cant threat to water shortage and a bigger challenge of water resources management when encountering the need to support a large amount of population. In a semi-arid environment, the water cycle has particular characteris-tics. Precipitation and many planetary boundary layer processes are a ected by soil moisture variability and, precisely in these areas, soil moisture is very variable in space and time, making the areas very vulnerable to climate change with consequences that may have serious social and environmental e ects (UNESCO, 2011). The water resources in semi-arid regions, with their vulnerability to climate change, need to be analysed, and the impact of climate change needs to be assessed.
The Mediterranean, as mentioned before, is one of the primary Hot-Spots of climate change. The precipitation is predicted to decrease as shown in Figure 1.1 (b), and droughts will occur more frequently. The Mediterranean region is often cited as an example of the extreme tension on water resources: 180 million inhabitants have less than 1000 m3/year per capita of renewable water resources (425 in Tunisia and 750 in Morocco), and 80 million will have less than 500 m3/year in 2025 (UNEP/MAP, 2009; EASAC , 2010). In most Mediterranean countries, overexploitation of water resources is already observed, and this situation is bound to deteriorate further. The water management is then, with its corollary of food security, the critical issue of the Mediterranean in the 21st century.
Water cycle and anthropogenic interventions
Water is not a stationary substance on Earth. It is moving all the time. To study the water resources, which form water cycle, their movements need to be known rst. The water cycle is the movement of water between the oceans, the atmosphere, the surface, and the fresh water. It consists of several processes including precipitation, evaporation, transpiration, in ltration, and runo , as shown in Figure 1.4 (Green, 2000).
Humans in uence the water cycle mainly by in uencing evapotranspiration (evaporation + transpiration) through irrigation for agriculture or land use change and runo through damming for water storage, hydro powers or ood control. The other processes are subsequently in u-enced. Anthropogenic in uences have a ected the global water cycle since the beginning of agriculture, and the impact has increased dramatically since the 1960s, as dams were built ex-tensively (Pachauri and Meyer, 2014).
With the construction of dams, reservoirs are usually formed from the water accumulating above them. Proper management of dams is of great importance since it relates not only to food production but also to oods, droughts, and energy. Moreover, the severe mismanaging of the dams may cause droughts in the downstream of the river or increase the risk of catastrophic oods. Several studies (Li et al., 2016; Haddeland et al., 2014; Gao et al., 2013; Guo et al., 2012) found that the construction of dams reduces river ows and changes the stream ow seasonally, which in uences the water cycle signi cantly. The impacts of reservoirs are most profound in Asia, Europe, and Africa, where in some months the total ux of fresh water into the ocean is 10% less compared to a naturalised situation (Biemans et al., 2011). Therefore, in addition to the satisfaction of the irrigation or energy purpose, the impact and role of dams in the water cycle need to be accounted for.
Equally important as dam construction, but still related to the dams, irrigation is another an-thropogenic intervention of the water cycle. To ful l the increasing global food demand, irrigation is increasing, making more contributions to evaporation as an important step in the water cycle. Irrigation creates an anthropogenic change of the soil moisture, and can ultimately in uence the planetary boundary layer, local circulation, clouds and precipitation through evaporation from soil and transpiration from thriving vegetation. Several studies (Tuinenburg et al., 2014; Puma and Cook, 2010; Lee et al., 2011; Douglas et al., 2009) found that irrigation increased evaporation both regionally and globally, and decreased temperature in the local area. Lo et al. (Lo and Famiglietti, 2013) identi ed that irrigation in the Central Valley (California, United States) increased summer precipitation by 15%, causing correspondingly the increase in Colorado River stream ow of about 30%. Boucher et al., 2004 (Boucher et al., 2004) estimated a global mean radiative forcing in the range of 0.03 to +0.1 W/m2 due to the increase in water vapour from irrigation, and a massive surface cooling of up to 0.8 K over irrigated land areas. Climate change with rising temperature globally increases atmospheric evaporation demand, and also, evaporation caused by irrigation in uences the climate. The water cycle involves the exchange of energy, which leads to temperature changes. The heat exchanges from water evaporation and condensation in uence climate and the impact need to be studied and involved in the water cycle process. Also, as the primary purpose of reservoirs, irrigation has been found to decrease the river discharge to the oceans both on a continental and global scale (Gerten et al., 2008; Biemans et al., 2011).
Land use changes also in uence the water cycle by changing the evapotranspiration. Land use changes may have many situations including agriculture to buildings, agriculture type changes, etc. With the complexity, it is di cult to quantify the in uences of land use change, and thus, out of the scope of this thesis.
Hydrological models and land surface models
In response to water-related challenges, many hydrological models have been developed to analyse, understand, and explore solutions for sustainable water management in the basin scale, to support decision makers and operational water managers. Hydrological models are mainly used for understanding hydrological processes, estimating the water resources, and predicting system behaviour. They have evolved from empirical rainfall-runo models (Mulvany, 1851; Sherman, 1932; Horton, 1935) to conceptual models (Linsley and Crawford, 1960; Crawford and Linsley, 1966; Dawdy and O’Donnell, 1965; Sugawara, 1969; Burnash et al., 1973; Burnash, 1995; Arnold et al., 1998; Ma and Cheng, 1998), and to physically based models (Abbott et al., 1986a,b; Beven and Kirkby, 1976, 1979; Beven et al., 1987; Grayson et al., 1992; Fortin et al., 2001a,b; Qu, 2004; Qu and Du y, 2007; Kumar, 2009; Shi et al., 2013; Todini, 2007). The recent physical models use parameters to characterise the unique aspects of the system being studied (Nepal et al., 2017). However, the parameters may di er in di erent catchments due to physical and hydrological di erences (Bloschl and Zehe, 2005).
In addition to hydrological models, Land Surface Models (LSMs) are equally important to learn the regional land surface status with more complexity since it is physical with the simulation of the energy balance. The LSMs are the components of Global Climate Models (GCMs) that simulate land surface processes, such as the absorption and partitioning of radiation, moisture, and carbon (Abramowitz et al., 2008). The LSMs model the surface energy balance, surface water balance and carbon balance. They are typically provided with meteorological conditions as inputs (from a boundary layer atmospheric model or reanalysis data) and produce outputs that include the surface runo , latent and sensible heat uxes, CO2 uxes, deep soil drainage, soil moisture, etc. LSMs are essential to assess the impact of climate change.
A model is a simpli ed representation of real-world system (Sharma et al., 2008), with the best one giving results close to reality with the use of least parameters and model complexity (Devia et al., 2015). However, most of the current models do not incorporate (or very little) human intervention in the water cycle, bringing more uncertainties into the assessment, analysis and forecasting. The uncertainties are even more signi cant in semi-arid areas since models are not accurately developed, and the impact of irrigation and reservoir operation may be ampli ed (Zhou et al., 2016).
Incorporating anthropogenic intervention in the models helps to understand the hydrological processes and the climate change better, yet long-term and consistent records that focus on water management and human interventions on the water cycle are not widely available (Zhou et al., 2016). The modelling requires better accuracy, both in time and spatial scale, for which researchers are still working on.
To solve water-related challenges and to assess the impact of climate change, better char-acterised hydrological models and LSMs are needed. For that, the anthropogenic intervention needs to be introduced in the water cycle and models. Involving anthropogenic intervention in the water cycle needs to be done by introducing the irrigation status and the water availability in rivers and reservoirs considering the e ects of dams.
LSMs can bene t from reliable soil moisture data for model validation and data assimila-tion. Soil moisture, which is directly linked with evaporation and decides the irrigation strategy, needs be to mapped with better e ciency to obtain a higher spatial and temporal resolution. Also, LSMs need a good mapping of irrigation, in order to have realistic covers and explicitly simulate irrigation, which is needed for cropland irrigation management (Ambika et al., 2016). The amount of irrigation surfaces, which is of great importance to water resources management, is still unclear. Therefore, to optimise the models, surfaces of irrigated and non-irrigated elds need to be mapped.
Furthermore, now that LSMs have started to include dams, they need dam level/volume data, which is often di cult to obtain. Temporal continuous water availability in rivers and reservoirs needs to be known to assess the e ects of dams in hydrological processes. Precise water levels over reservoirs should be monitored to evaluate the available water amount for irrigations. The bu ering e ects of dams should be included and assessed in the models for optimisation.
Measurements and remote sensing techniques
The irrigation status is traditionally measured by eld campaign, which is time-consuming and more importantly impossible to be done for all areas. Soil moisture is traditionally measured with in situ measurements either by taking soil samples or by the soil moisture probe. However, in situ networks represent single point locations and usually cover relatively short periods of observation (Pratola et al., 2015).
Considering the aspect of water levels, the traditional way of monitoring is using the gauging station or stream gauge, which is a location used by hydrologists to monitor the terrestrial bodies of water. Unfortunately, in situ gauging stations are not always available in many parts of the world, or they are otherwise not publicly available and are maintained by local authorities. Furthermore, over the last few decades, in situ observations of hydrologic variables have generally been in decline (Trenberth et al., 2018).
The access to both data sources is critical, and moreover, continuous temporal data is needed for the monitoring, assessing, modelling and predicting. Yet data are sparse in developing coun-tries both in spatial and in temporal dimensions. Even where such data have been collected, they are rarely shared across ministries or institutions, while thousands of freshwater bodies are the shared resource of several countries.
At the same time, new observation methods that use satellite sensors have thrived over the past few decades. Thanks to the remote sensing technology, we now possess a powerful tool for monitoring the water resources both within inland water systems where no in situ data are available and within transboundary river basins. Wherever satellite data are available, the surface soil moisture, the status of irrigation, and the surface water levels can be assessed and monitored at both regional and global scales.
Microwave remote sensing
In particular, microwave remote sensing uses electromagnetic radiation with a wavelength be-tween 1 cm and 1 m (300 MHz to 30 GHz commonly referred to as microwaves) as a measurement tool. While electromagnetic waves in the visible and infrared portions of the spectrum primarily characterised by wavelength, microwave portions of the spectrum are often referenced according to both wavelength and frequency (Natural Resources Canada, 2015). The microwave region of the spectrum is quite large, relative to the visible and infrared, and there are several wavelength ranges or bands commonly used which given code letters during World War II, and remained to today as listed in Table 2.1 (du Preez and Sinha, 2016).
Microwave remote sensing has many advantages compared to optical remote sensing. It has a day-and-night imaging capability, and ability to penetrate could cover, haze, fog, and to some extent, rain as shown in Figure 2.3. For lower frequencies at L- and P-band, the transmission rate is almost 100%. Moreover, microwave remote sensing is more sensitive to the dielectric constant of the medium and structural characteristics of the surface, such as roughness or canopy structure compared to optical sensors (Ulaby et al., 1986a; Dobson and Ulaby, 1981).
Ranging system and principle of altimetry
Non-imaging radars, including altimeters, measure the elevation of the earth’s surface, and scatterometers are used to estimate ocean wind speed. In this section, the altimeter, which is the ranging system, is discussed.
Space-borne radar altimeters are essential tools for monitoring the oceans, which has been performed for over 20 years (Benveniste, 2011). Satellite altimeters have also proven to be valuable tools for monitoring the water levels within inland water systems, including lakes and rivers (Alsdorf et al., 2007; Calman and Seyler, 2006; Cretaux and Birkett, 2006).
Space-borne radar altimeters transmit a short microwave pulse in the nadir direction, and the signal re ected by the surface is received by the instruments. As shown in Figure 2.9, the altimeter measures the time delay taken for a radar pulse to travel to the surface and back again. The elapsed time corresponds to the range between the satellite and the Earth’s surface (Fernandes et al., 2014) by the following equation: R = c ∗ T /2 (2.12).
Table of contents :
1.1.1 Climate change and interaction with water resources
1.1.2 Water resources in global scale
1.1.3 Water resources in regional scale
1.1.4 Water cycle and anthropogenic interventions
1.2 Water resources estimation
1.2.1 Hydrological models and land surface models
1.2.2 Measurements and remote sensing techniques
1.3 About the thesis
1.3.2 Structure of the thesis
2 Remote Sensing Principles
2.1.1 Electromagnetic waves
2.1.2 Characteristics of electromagnetic waves
2.2 Microwave remote sensing
2.2.2 Synthetic Aperture Radar
2.2.3 Imaging systems
184.108.40.206 Scattering mechanisms
220.127.116.11 Topographic eects
18.104.22.168 Speckle eect
22.214.171.124 Soil moisture principle
2.2.4 Ranging system and principle of altimetry
126.96.36.199 Trackers and Retrackers
188.8.131.52 Pulse-limited altimetry
184.108.40.206 SAR/Delay-Doppler altimetry
3 Study Area and Datasets
3.1 Study area
3.1.1 Ebro basin
3.1.2 Urgell site
3.2 Remote sensing dataset
3.2.1 Sentinel data
3.2.2 MODIS data
3.2.3 DEM data
3.2.4 SMOS data
3.3 In situ data
3.3.1 Ground soil moisture measurements
3.3.2 Meteorological data
3.3.4 SAIH Ebro
4 Soil Moisture Retrieval
4.2.1 Backscatter dierence with the driest date
4.2.2 Backscatter dierence between two consecutive dates
4.3 Soil moisture at 1 km resolution
4.4 Soil moisture at 100 m resolution
4.5 Published paper
4.6 Summary and Conclusions
5 Irrigation Mapping
5.3 Published paper
5.4 Summary and Conclusions
6 Water Level Retrieval
6.3 Published paper
6.4 Summary and Conclusions
7 Dam Simulation
7.2.1 SURFEX land surface model
7.2.2 SAFRAN meteorological analysis system
7.2.3 SASER hydrological model
7.3.1 Forcing model
7.3.2 Conversion from water level to volume
7.4.2 Satellite data forcing
8 Conclusions and Perspectives
8.1 Main conclusions
8.2 Future research lines
List of Figures
List of Tables