Hydrological modelling improvements in the assessment of water resources of agrarian subbasins in semiarid regions

Resumen

Hydrological modelling is nowadays an essential step for sustainable management of water and land resources. Despite the fact that most of the hydrological processes are well known and equations performance is accurate, most of the challenge in hydrological modelling focusing on whether the model can, in some close way, represent the real world. Most hydrological models are generally biased during modelling exercises, including setting up and interpreting the results. The correct representation depends on the input data, which also depend on two practices: i) the correct harmonization of the data and the available scales, before the configuration of the model and ii) the use of adequate tools and complementary for data pre-processing. The absence of these practices leads to a reduction in the quality of the data and therefore a decrease in the precision of the results. The focus of this thesis is in the improvements about hydrological modelling in agrarian sub-basins in the semiarid. These environments are especially sensitive to biophysical aspects as climate, soils, vegetation and land management. This work offers a complementary approach to hydrological modelling of agrarian watersheds in Spain, through using the Soil Water and Assessment Tool (SWAT). In Spain, when regional modelling and long-term modelling is required, model inputs data are scarce and scales are not usually compatible. One of the main, untreated inputs, to run models over these characteristics is the soil data. In this work, the Self-Organizing maps (SOM) are presented as an alternative method to improve digital soil mapping for hydrological modelling getting promising improvements in comparison to taxonomic classic approach. There is not previous evidence in literature using this method for soil mapping in hydrology modelling. Other sensitive data is the land use land cover (LULC), as a spatial dynamic array that influences water flows. Using Earth Observation (EO) and surveys allow including a large dataset of crop rotations schemas and crop practices to model more realistic agrarian effect at subbasin scale. Once that model is calibrated and validated, LULC scenarios can be assessed to determine the influence of future land policy making in water resources. The former model improvements were developed in one of the most recurrent drought alerted watershed in last decades in Duero River basin, the Cega-Eresma-Adaja (CEA) exploitation system. An exploitation system is referred as a management system grouping watersheds with similar characteristics (i.e. biophysical parameters, climate and land management) but this grouping is done to facilitate management and decision making by the River Basin Authority. However, there are important differences from technical point of view referred to hydrology and land and water dynamics for each River implies different water balance flows as demonstrated in this work. In this work, was studied the influence of different soil maps and resolutions on the main hydrological components of a sub-arid watershed. The Soil Water and Assessment Tool (SWAT) was parameterized with three different soil maps. A first one was based on Harmonized World Soil database from FAO, at scale 1:1,000,000 (HWSD). The other two were based on a Kriging interpolation at 100x100 m from soil samples. To obtain soil properties map from it, two strategies were applied: one was to average the soil properties following the official taxonomic soil units at 1:400,000 scale (Agricultural Technological Institute of Castilla and Leon - ITACyL) and the other was to applied Self-organizing map (SOM) to create the soil units (SOMM). The results suggest that scale and soil properties mapping influence HRU definition, which in turn affects water flow through the soils. Statistical metrics of model performance were improved from R2 =0.62 and NSE=0.46 with HWSD soil map to R2 =0.86 and NSE=0.84 with SOM and similar values were achieved during validation. The CEA watershed for the period 2004-2014 was calibrated and validated analysing hydrological year types to provide more details of low-flows during spring-summer periods. The study reveals that aspects such as crop rotation, soil management and their associated measures in Mediterranean basins are key factors for water resource management facing climate change. These results are expected to serve stakeholders and River Basin Authorities in conducting better-integrated water management practices in the watershed. Cereals in CEA midlands are a predominant crop choice because of climate and soil factors. These crops represent most of the agricultural water demand in the midlands. Crop rotations of wheat, fallow and barley are major choices within cereal crop sequences. Characterizing agricultural land processes coupling weather and soils are challenging because of multeity of factors affecting vegetation growth of cereals. One of these growth factors in semiarid is especially the rainfall on agricultural fields, in which soil properties and climate are strongly correlated with crop yield. These relationships are commonly analysed using vegetation indices such as the normalized difference vegetation index (NDVI). NDVI series from two zones, belonging to different agroclimatic zones from CEA, were examined decomposing them into the overall average pattern, the residual series, and anomalies series. All of them studied by applying the concept of the generalized Hurst exponent (GHE). This is derived from the generalised structure function (GSF), which characterizes the series’ scaling properties. The overall pattern of the NDVI original series, NVDI residual and NDVI anomalies were examined from both zones. These presenting differences explained from climate-soil characteristics. The significant differences found in the soil reflectance bands confirm the differences in these two zones. Original NDVI series are persistent and multiscaling as other works reports. With respect to the scaling properties of the NDVI residual series, these presented Hurst exponents significantly lower than 0.5 indicating the structure of the signals. A stronger anti-persistent character was obtained in NDVI residual series with significant differences between zones. Similar is the case of NDVI anomalies with minor scaling properties. These findings reveal the influences of soil-climate interactions in the dynamic of the NDVI series for rainfed cereal crops in a semiarid climate. The assessment of land use and land cover (LULC) scenarios is a relevant field of study to anticipate future environmental impacts at the basin scale. Often the LULC scenarios and transition rules for hydrological modelling are based on expert criteria and disregard a participatory approach for its definition. In this work was analyzed the potential implications of three stakeholder informed LULC scenarios, and its implications in the water balance components of a sub-arid catchment of CEA. The LULC scenarios were defined through a participatory scenario process, involving a wide range of stakeholders and experts, and reflect three contrasting local land use developments: Land Sharing (LSH), Land sparing (LSP) and Land balance (LBA). The SWAT model was used to quantify the water resources implications linked to the LULC baseline scenario and the alternative LULC futures. The three scenario narratives underpinning the modelling scenarios highlight differences among reforestation transitions, landscape fragmentation, cropping patterns, and agricultural specialization.