TL;DR
This paper introduces a new, more realistic irrigation module into the ORCHIDEE-CROP land surface model, allowing for simulations of different irrigation techniques and their impact on China's water resources. The improved model successfully captures broad patterns of irrigation water use and highlights irrigation efficiency and land use change as key drivers of irrigation trends.
What Problem Does This Paper Address?
Accurately modeling the global water cycle requires a robust representation of human water management, particularly irrigation, which accounts for a significant portion of global freshwater use and food production. Previous land surface models often simplified irrigation, treating crops uniformly or failing to account for specific techniques like flood or paddy irrigation. This simplification limits our ability to understand how irrigation impacts regional water budgets, energy fluxes, and ultimately, climate. This study addresses this gap by developing and implementing a more sophisticated irrigation module within the ORCHIDEE-CROP model, specifically tailored to capture the complexities of irrigation practices in China, a region with widespread and diverse irrigation systems.
Methods
The core of this work is the development of a new irrigation module for the ORCHIDEE-CROP land surface model. This module incorporates three key irrigation techniques: drip, flood, and paddy irrigation, each with distinct water application and water balance considerations. The model simulates irrigation demand based on crop water stress and then attempts to meet this demand using available surface water resources (streamflow, surface runoff, and deep drainage). Crucially, the model also includes an "unlimited irrigation" (FI) scenario to represent conditions where irrigation demand is met, even if it exceeds available renewable water resources, which is useful for estimating potential crop yields and understanding irrigation in water-scarce regions where groundwater might be heavily utilized. The study then performs three simulations: no irrigation (NI), irrigation limited by local water resources (IR), and full irrigation (FI), to analyze the impacts of irrigation.
Datasets
The study utilizes several key datasets:
- China 15-PFT Map: A land cover map categorizing vegetation into 15 Plant Functional Types, with a focus on four crop types (wheat, maize, rice, and other crops) for China. This map was updated with temporal changes from 1982 to 2016 based on provincial census data.
- Global Map of Irrigation Area (GMIA) version 5: Used to derive the irrigation fraction (f) for each grid cell.
- Provincial Irrigation Withdrawal Data (Wstat): Statistical data from 1990-2013 used for validating simulated irrigation withdrawal.
- GSWP3 Atmospheric Forcing Data: Used to drive the model simulations, providing 3-hourly temporal and 0.5° spatial resolution data.
- GRACE Total Water Storage (TWS) Products: Used to evaluate the model's simulation of water storage changes.
- Three Gorges Dam (TGD) Water Storage Data: Used to correct for the influence of artificial reservoirs on TWS seasonality in the Yangtze River Basin.
Key Insights and Ideas
- Irrigation Techniques Matter: Explicitly modeling different irrigation techniques (drip, flood, paddy) is crucial for accurately representing water demand and application, especially in regions like China where paddy rice is significant.
- Water Limitation is Key: The difference between the "limited by local water resources" (IR) and "full irrigation" (FI) scenarios highlights the significant impact of water availability on actual irrigation withdrawal, particularly in water-scarce regions like the North China Plain.
- Irrigation Efficiency Drives Trends: While climate change influences irrigation demand, improvements in irrigation efficiency appear to be the dominant factor behind the observed decrease in overall irrigation water withdrawal in China over recent decades.
- Land Use Change is Localized: Land use change (e.g., cropland expansion or contraction) can significantly influence irrigation demand at a local or regional scale, but its effect on total national water withdrawal is less pronounced due to broader water resource limitations.
- Model Captures TWS Seasonality, but Not All Drivers: The model successfully simulates the seasonal variations in total water storage (TWS) driven by precipitation and evapotranspiration, but it struggles to fully capture the impact of human interventions like reservoir regulation and groundwater depletion without specific modules for these processes.
Key Results
- The simulated irrigation water withdrawal (W) from the IR and FI scenarios showed fair spatial agreement with statistical data (r = 0.68 ± 0.07).
- The FI simulation generally overestimated water withdrawal in water-limited regions compared to the IR simulation, which was better constrained by available water.
- The model successfully reproduced the seasonal variations of Total Water Storage (TWS) when compared to GRACE satellite data (r = 0.87).
- Irrigation was found to increase evapotranspiration (ET) by 11.6% (IR) and 25.7% (FI) nationally.
- Irrigation led to a slight decrease in mean surface temperature across China, with a more pronounced cooling effect in the North China Plain.
- The model's simulated TWS trends aligned well with GRACE data, but the decreasing trend in the Yellow River Basin due to groundwater depletion was not fully captured without inferring groundwater withdrawal.
Limitations and Future Work
- Groundwater and Water Transfer: The model's inability to explicitly simulate groundwater extraction and water transfer mechanisms leads to underestimation of irrigation withdrawal in arid and semi-arid regions where these sources are critical. Future work should incorporate these processes.
- Irrigation Fraction Data: The use of a static irrigation fraction map is a simplification. More dynamic and accurate data on irrigated area trends are needed.
- Irrigation Efficiency Data: While irrigation efficiency is shown to be important, precise gridded data on its spatial variations and trends are lacking.
- Atmospheric Coupling: The current offline simulation of surface energy budget might overestimate the cooling effect of irrigation due to the lack of atmospheric feedback. Future work should involve coupling the land surface model with an atmospheric model.
- Artificial Reservoirs and Groundwater: The study acknowledges that the seasonality of TWS in the Yangtze River Basin is strongly influenced by artificial reservoirs, and the decreasing trend in the Yellow River Basin is driven by groundwater depletion. Including these human impacts in future simulations is crucial for more accurate water resource projections.
- Other Crops and Multicropping: The model currently focuses on wheat, maize, and rice. Future work could expand to include other crops and account for multicropping practices, which are common in some regions.
Summary updated