Changes in terrestrial water stress and contributions of major factors under temperature rise constraint scenarios

Abstract

The Paris agreement adopted at the 21st Conference of Parties of the United Nations Framework Convention on Climate Change stipulates 2 and 1.5 °C targets, but their consistency with sustainable development is poorly understood. This study focuses on water stress defined by annual water consumption-to-availability ratio (CAR) and analyzes the CAR changes for 32 global regions during this century for scenarios of the 2 and 1.5 °C targets. It also estimates contributions of major factors behind such change for addressing the adaptation planning. The results show that the CARs in many (i.e., 25) regions remain very small (less than 0.1) regardless of the future temperature level. For the other seven regions, the CARs undergo significant changes, while the changes and contributing factors to them are different by region and the future temperature level. Possible adaptation strategies are given for regions of significantly increasing CARs. For instance, in Afghanistan and Pakistan and South Africa, the CARs increase mainly due to increases in irrigation water associated with socioeconomic development (i.e., food demand growth). Decreases in water availability and increases in irrigation water due to climate change also contribute to the CAR increases after 2030. The contributions of other factors (i.e., demand changes in municipal water, water for electricity generation, other industrial water, and water for livestock) are small. In these regions, securing water resources as well as irrigation water conservation are important to avoid worsening of the CAR. Adaptation strategy recommendations for other regions are also presented.

Appendices

Appendices

1.1 Appendix 1 Formulation of the CAR change rate

The CAR is denoted using Eqs. (A1) and (A2).

$$ CAR= Use/R $$

(A1)

$$ Use=M+\left(E+ OI\right)+\left( Irri+L\right) $$

(A2)

• M: Municipal water

• E: Water for electricity generation

• OI: Other industrial water

• Irri: Irrigation water

• L: Water for livestock

• R: Renewable water availability (renewable surface and ground water)

Therefore, the change rate of CAR (∆CAR) for the period between t₁ and t₂ is denoted using Eqs. (A3) and (A4).

$$ \Delta CAR=\Delta Use-\Delta R $$

(A3)

$$ \Delta X=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{X_2}{X_1}\right) $$

(A4)

Furthermore, we decompose the ∆Use using approximate expressions denoted as Eqs. (A5) and (A6), to understand the individual contributions of municipal water, water for electricity generation, other industrial water, irrigation water, and water for livestock.

$$ \Delta Use\approx {\Delta}^{\prime }M+{\Delta}^{\prime } OI+{\Delta}^{\prime }E+{\Delta}^{\prime } Irri+{\Delta}^{\prime }L $$

(A5)

$$ \left.\begin{array}{l}{\Delta}^{\prime }M=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{M_2+{OI}_1+{E}_1+{Irri}_1+{L}_1}{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_1}\right)\\ {}{\Delta}^{\prime } OI=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{M_1+{OI}_2+{E}_1+{Irri}_1+{L}_1}{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_1}\right)\\ {}{\Delta}^{\prime }E=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{M_1+{OI}_1+{E}_2+{Irri}_1+{L}_1}{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_1}\right)\\ {}{\Delta}^{\prime } Irri=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{M_1+{OI}_1+{E}_1+{Irri}_2+{L}_1}{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_1}\right)\\ {}{\Delta}^{\prime }L=\frac{1}{t_2-{t}_1}\cdot \ln \left(\frac{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_2}{M_1+{OI}_1+{E}_1+{Irri}_1+{L}_1}\right)\end{array}\right\} $$

(A6)

The approximation was applied to all regions and cases, confirming that the approximation error for ΔUse is small enough not to affect the evaluation results.⁶

1.2 Appendix 2 Regional divisions in this study

Fig. 7

The locations of the 32 regions

Table 2 Correspondence between the 32 regions and the 6 zones

1.3 Appendix 3 Scenarios on socioeconomic development and food demand

The amounts of the 32 regions are aggregated into the six zones; for the six zones, refer to Table 2.

Fig. 8

Population scenario. Scenarios on urban and rural population are utilized for estimations of the municipal water. The ratio of urban population relative to total population is projected based on GDP of the each country (Hayashi et al. 2013)

Fig. 9

Per capita GDP scenario

1.4 Appendix 4 Supplements for estimation of water use in the power sector

Fig. 10

Relationships between thermal conversion efficiency and heat to cooling (HCE). Estimations based on Martín (2012)

Table 3 Water requirement per unit of energy rejected (WR)

Fig. 11

Relationships between thermal conversion efficiency and water for CO₂ separation and capture. Circles show examples for plants in Japan (Nakagami et al. 2016). The lines for oil and biomass are estimated based on the CO₂ emissions per calorie for each fuel (Garg et al. 2006)

Table 4 Water for other uses (for boilers, gasification etc.)

Table 5 A global mean of the thermal conversion efficiency estimated in the Reference and the no climate change cases (unit: %HHV)

1.5 Appendix 5 Supplements for estimation of water use in other industrial sectors

Table 6 The production volumes for representative manufacturing sectors (i.e., iron and steel, the chemical industry, and pulp and paper sectors) and the energy-use efficiency for the production of crude steel by blast furnaces. The values in 2000 and the annual change rates are listed for the top 20 water withdrawal regions for the other industrial activities in 2050 among the 54 regions defined for the DNE21+ model, which are expected to occupy over 75% of the world water withdrawal for the activities during the twenty-first century

1.6 Appendix 6 Supplements for estimation of the municipal water

Fig. 12

a, b Relationships between GDP and the ratio of the “PA” relative to the population for the period between 1990 and 2014 in developing countries. Data source: World Bank (2016). The “PA” means a person who is able to access safe water. Figures for developed countries are omitted, since the ratios for the countries have already reached around 100% in both urban and rural areas

The water withdrawal per PA is estimated by country based on Eq. (A7).

$$ \left.\begin{array}{l}\mathrm{DW}=k\times \mathrm{DW}0\\ {}\log \mathrm{DW}0=\left(a\times \mathrm{per}\;\mathrm{capita}\;\mathrm{GDP}\right)/\left(b+\mathrm{per}\;\mathrm{capita}\;\mathrm{GDP}\right)\end{array}\right\} $$

(A7)

where DW is per PA withdrawal, a and b are regression coefficients (a = 2.16 and b = 90 for urban areas, and a = 1.08 and b = 148 for rural areas), and k is a factor to reduce discrepancies between the per PA withdrawal estimated by the regression function and the amount by statistics for the year of 2000 (World Bank 2016).

1.7 Appendix 7 Supplements for estimation of the irrigation water

Table 7 Irrigated area estimated in the no climate change case (unit: 10³ km²). The values for each of the 32 regions defined for the GLaW model are listed in order of the irrigated area in 2050. Regions whose irrigated area is smaller than 20,000 km² are aggregated into “other regions”

1.8 Appendix 8 Water consumption and availability for the six zones

The 32 regions are aggregated into six zones; for the six zones, refer to Table 3.

Table 8 Water consumption and availability for the six zones (unit: km³ year⁻¹)

1.9 Appendix 9 The CAR in 2050 and 2100

Table 9 Numerical values of the CAR shown in Fig. 4