1. Introduction
Vegetation covers 70% of the Earth's surface and is a crucial ecosystem component \cite{Zhang_2019,Arora_2002}. Carbon capture by vegetation is considered one of the most effective strategies for limiting the increase of global CO2 concentrations. As an essential pathway for nature-based solutions, vegetation restoration has been proven effective in absorbing carbon dioxide \cite{Fang_2023,Robison_2023}, preventing soil erosion\cite{Hu_2023,Luo_2023}, and maintaining biodiversity\cite{Zhou_2023,Porto_2009}. However, large-scale vegetation restoration also brings some problems. It was found that revegetation consumes more surface water resources and reduces runoff, which may create potential conflicts with human water use\cite{Feng_2016,Jian_2015,Chang_2022}. These adverse effects are particularly severe in arid and semi-arid regions\cite{Bai_2020,Zhou_2020}.
Drylands (defined as areas with arid index <0.65) \cite{Pan_2021}, with an approximate area of 5240000 km2 in China, are the typical ecologically fragile region \cite{Wang_2023,Huang_2015}. Since the late 1970s, the Chinese government has implemented many vegetation restoration measures here to protect the ecological environment and restore degraded ecosystems, including the Grain for Green (GFG) project, the Three-North Shelter Forest Program, ecological restoration projects, etc. This series of projects has led to a significant greening trend in drylands \cite{Chen_2019,Liu_2012}. However, the growth of large amounts of vegetation consumes many water resources through evapotranspiration \cite{Anderson_2012,Cao_2011}. This reduced local runoff and caused soil desiccation in deep soil layers \cite{Liang_2015}, even exerting excessive pressure on terrestrial water reserves, further threatening the sustainability of fragile ecosystems \cite{Zhao_2020}. Meanwhile, as socio-economic development and population increase, human water demand also increases day by day \cite{Huang_2018}. So, in the background of insufficient water supply and increasing demand, how to fully utilize limited water resources for vegetation restoration in drylands has become a crucial issue that needs to be answered urgently.
Previous studies have attempted to find the upper limit of vegetation-carrying capacity to contribute to sustainable development. Some studies construct relationships between vegetation and local climatic or geomorphological conditions (e.g., topography, temperature, precipitation, etc.) to calculate vegetation carrying capacity \cite{Zhang_2019a,Nauman_2017}. For example, Zhang et al. \cite{Zhang_2020} defined the carrying capacity for vegetation as the vegetation threshold under abiotic stresses (temperature, precipitation, potential evapotranspiration, and elevation), reflecting the multi-year growth of vegetation. This method is highly dependent on the surrounding environment, and does not take into account, or only partially takes into account, water resources, so the vegetation restoration potential obtained may exceed the water-carrying capacity. Some other studies use model simulations, such as vegetation dynamic models (黄土高原潜在自然植被空间格局及其生境适宜性10.13870/j.cnki.stbcxb.2021.05.026), ecological hydrological models \cite{Zhang_2018}, biogeochemical models \cite{Jia_2019}, to calculate vegetation growth limitations. However, the simulation accuracy is still being determined due to the difficulty of quantifying some parameters. Other studies calculate vegetation restoration thresholds based on water balance \cite{Li_2023,WANG_Kaili_2022}. Feng et al. \cite{Feng_2016} constructed a quantitative relationship between ET and NPP in the Loess Plateau, taking into account the effect of human water use, but the study did not provide the spatial variation of vegetation restoration potential. On this basis, Liang et al. \cite{Liang_2019} constructed a fitted relationship between GPP and ET under different vegetation, topography, and precipitation conditions to explore the vegetation restoration thresholds under current and future water resource scenarios. These past studies mainly assessed the overall vegetation condition without distinguishing vegetation types, and the focus on the drylands of China is still insufficient.
In this study, additional water resources (AWR) available for vegetation restoration were defined as precipitation minus natural water consumption and human water use. Our study aims to explore (1) how much AWR is currently available for vegetation restoration, (2) based on the current water resources and land cover, how much revegetation potential is available for different vegetation in drylands of China, and (3) how much vegetation conversion we could do under water-limitation. We used multiple data combinations to minimize errors introduced by the data, and the results are presented as medians.
2. Methods and data
2.1 Data
Precipitation (P) was obtained from the following multiple datasets: (1) the China Meteorological Forcing Dataset (CMFD), which was developed by the Institute of Tibetan Plateau Research under the Chinese Academy of Sciences, with 0.1° spatial resolution and monthly temporal resolution from 1979 to 2018
(). (2) the CN05 dataset, derived from interpolating observations from more than 2400 stations throughout mainland China, with 0.25° spatial resolution and monthly temporal resolution from 1961 to 2021
(10.6038/cjg20130406). (3) the ERA5 reanalysis dataset from the European Center for Medium-Range Weather Forecasts (ECMWF), which provided monthly precipitation from 1950 to 2022 at a 0.1° spatial resolution (
https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5).
Evapotranspiration (ET) was also estimated based on multiple datasets, including (1) the Global Land Evaporation (GLEAM v3.6a), which was a combination of gauge-based, reanalysis, and satellite-based data, spanning the 42 years 1980-2021 at a spatial resolution of 0.25° (
https://www.gleam.eu/). (2) PML_V2 dataset, coupling with the GPP process, at 8-day intervals from 2002 to 2019 and a spatial resolution of 0.05°
(10.11888/Geogra.tpdc.270251). (3) ET from Song et al.
(?), which gave monthly ET from 2001 to 2020 at 0.01° spatial resolution.
Human water consumption data was derived from a previous study \cite{Huang_2018}. It provides a global monthly gridded (0.5°) sectoral water use dataset from 1971-2010. To match with other data, we referred to the spatial pattern of water use in that study and combined it with the Water Resources Bulletin to extend human water use data to 2018.
Sources of other associated data are as follows: GPP was also derived from PML_V2 dataset. Irrigation area data of China from Zhang et al.
\cite{Zhang_2022}. The land cover map of the drylands in 2015 with a spatial resolution of 1 km was obtained from Resource and Environment Science and Data Center (
https://www.resdc.cn/DOI/DOI.aspx?DOIID=54).
Given the multiple precipitation and ET datasets, there are nine data combinations for calculating additional water availability and three combinations for calculating WUE. We used the ensemble median values to characterize the results and analyze the uncertainty.
2.2 Methods
2.2.1 Calculation of additional water resources available for vegetation restoration (AWR)
As shown in Fig.\ref{990698}, we first calculated the extra water resources that could be used for vegetation restoration in 0.5-degree grids. Since surface water is generally considered to be precipitation minus evapotranspiration \cite{Zhou_2022,Rockstr_m_2009}, we define additional water available for revegetation (AWR) as follows: