Towards a Dynamic Irrigation Representation in Land Surface Models
by Wanxue Zhu
Date of Examination:2025-10-16
Date of issue:2025-10-29
Advisor:Prof. Dr. Stefan Siebert
Referee:Prof. Dr. Stefan Siebert
Referee:Prof. Dr. Thomas Heckelei
Persistent Address:
http://dx.doi.org/10.53846/goediss-11603
Files in this item
Name:Thesis_wanxue_eDiss_removeCV_20251024.pdf
Size:11.4Mb
Format:PDF
Description:PhD thesis-Wanxue Zhu
The following license files are associated with this item:
Abstract
English
Irrigation is crucial for sustaining agricultural productivity and food security, accounting for over 70% of global freshwater withdrawals. Beyond stabilizing crop yields, it increasingly serves as a key climate adaptation strategy in regions facing intensified hydroclimatic stress. By supplementing rainfall, irrigation supports crop growth in marginal areas, modulates land surface energy balance, and buffers agricultural systems against climatic variability. Although traditionally linked to arid and semi-arid zones, irrigation has gained importance in Europe, where climatic conditions range from drought-prone Mediterranean regions to temperate, humid zones. Climate-driven changes such as altered precipitation patterns, more frequent droughts, and intensified heatwaves are reshaping crop water demand and irrigation practices. Consequently, understanding irrigation dynamics in Europe is essential to guide sustainable water management and enhance agricultural resilience across diverse climatic conditions. The volume of irrigation water use is strongly influenced by the extent of irrigated area. Despite its growing importance, long-term and spatially explicit datasets capturing irrigation area dynamics across Europe remain limited, particularly those disaggregated by crop type. This persistent data gap constrains the assessment of agricultural water demand, the monitoring of adaptation responses, and the design of evidence-based water governance. To address this challenge, this dissertation constructs a suite of consistent, climate-driven, and crop-specific irrigation area datasets covering multiple decades by integrating process-based modeling with multi-source data and spatial disaggregation strategies. The central hypothesis is that drought conditions exert a dominant influence on the actual use of irrigation infrastructure, such that interannual variability in irrigation area can be inferred from drought signals. This dissertation comprises three interconnected studies that advance the understanding of irrigation dynamics across Europe. Section 2 examines interannual variability in irrigation infrastructure use in response to drought conditions, leading to the reconstruction of the European Long-term Irrigation Area Dataset (ELIAD). Specifically, it reveals that drought tend to stimulate irrigation practices in humid regions such as Western Europe and Northern Europe, whereas in arid areas like Spain in Southern Europe, water scarcity constrains irrigation, resulting in a negative correlation between drought and irrigated area. Historical and socio-economic factors also moderate this relationship, producing weaker responses in Eastern Europe and parts of Central Europe. Furthermore, this study describes the development of the first annual dataset of total irrigable and irrigated areas for 32 European countries from 1990 to 2020, identifying peak irrigation extents during drought years (2003 and 2018, with 11.93 and 11.77 million hectares, respectively) and minima during wet years (2002 and 2014, with 10.71 and 10.31 million hectares). It also analyzes the spatial heterogeneity of irrigation patterns between dry and wet years, revealing distinct regional responses. By characterizing the long-term spatial and temporal variability of irrigation, this work provides a crucial baseline for assessing agricultural water demand and guiding adaptive management under climate variability. Following this, section 3 describes the development and validation of the European Crop-specific Irrigated Area (ECIRA) dataset. It provides annual growing areas of 16 major irrigated and rainfed crops a 1-km resolution, for 28 European countries from 2010 to 2020. ECIRA disaggregates the subnational total irrigation areas from ELIAD by integrating statistical survey data and gridded information on crop distribution and irrigation infrastructure. It demonstrates high consistency with national census data, LUCAS field surveys, and existing irrigation datasets (i.e., MIRCA-OS, SPAM, and EIM2010), while substantially improving spatial resolution (1 km vs. ~10 km) to better characterize crop-specific irrigation patterns. Despite minor discrepancies in humid regions or for crops with low irrigation prevalence, ECIRA effectively captures spatial hotspots of major irrigated crops, offering a reliable basis for agricultural and hydrological applications across Europe. This enhanced spatial and thematic detail provides a robust foundation for crop water modeling, land surface simulations, and policy-relevant assessments of irrigation. Section 4 of this thesis presents a comprehensive review of the state of large-scale irrigation area mapping. It first reviews the properties of major global and regional (Europe, India, China, and the United States) irrigation data sets, and highlights their methodological linkages, strengths, and limitations. Using Europe as a case study, it then benchmarks existing datasets alongside the developed ELIAD and ECIRA products, against EUROSTAT 2020 gridded statistics. The evaluation includes visual consistency checks, national-level aggregation comparisons, and 10 km pixel-wise regression analyses. Results show that in humid regions, where irrigation is often supplementary and produces weak spectral and temporal signals, methods supported by ground-truth information (e.g., field and statistical surveys) achieve substantially higher accuracy, whereas datasets lacking such information tend to be less reliable (R² = 0.7 vs. 0.2). Next, the review explores how current irrigation mapping approaches leverage spectral, temporal, and spatial features to identify and characterize irrigation practices. Lastly, it outlines key priorities for future research, emphasizing the role of multi-source data integration and ground-based observations in advancing robust irrigation mapping frameworks. Collectively, this review clarify the current landscape of irrigation data and mapping methodology, and provide a roadmap for advancing data quality and usability in future research and applications. Sections 2–4 are based on three articles published in (sections 2 and 3) or submitted to (section 4) highly ranked scientific journals. The overall contributions of this thesis can be summarized in three key aspects: (1) irrigation responses to climate variability are elucidated, highlighting contrasting effects of drought stimulation across European climatic zones; (2) an important data gap is addressed through the construction of long-term, high-resolution, crop-specific irrigation datasets for Europe, supported by a scalable framework integrating process-based modeling with multi-source data and spatial disaggregation; (3) existing large-scale irrigation area products are systematically reviewed and evaluated, clarifying their interconnections, strengths, and limitations, thereby deepening the understanding of irrigation mapping principles and providing guidance for future improvements in accuracy and applicability.
Keywords: Irrigation; Climate variability; Crop-specific mapping; Data integration; Europe
