Case study: Extending the PRINCE water use results with water scarcity weighting

Water extraction in water-rich and water-scarce areas can have very different impacts on ecosystems and the services they offer local people. This case study describes work done under the PRINCE project to explore ways of reflecting water scarcity in the basic PRINCE water use macro-indicators.

Download as PDF

Background

The water use statistics in the PRINCE accounts provide information on the volume of water use (in Mm3) embedded in goods and services consumed in Sweden. But they are not calculated in a way that can reflect the differentiated impacts extraction of water is likely to have on ecosystem services in regions of production. In order to assess the potential localised socio-environmental effects of water extraction, water scarcity (i.e. the relative availability of sustainable freshwater resources) is an important consideration. In this case study, two methods are used to connect indices of water scarcity to the PRINCE water consumption outputs.

Theory and methods

Water scarcity is often presented as a ratio between demand and supply (a water-to-availability, WTA, ratio), which in turn can be assessed against scarcity thresholds – for example, moderate water stress has been previously defined as a WTA ratio above 0.2 and severe water stress as a WTA ratio above 0.4 (Vorosmarty 2000) – and thus converted into water scarcity indices (see e.g. Pfister et al. 2009).

The present case study adopts two methods for linking water scarcity indices (WSIs) with PRINCE data. The first method relies on water scarcity data compiled for the EU-FP7 CREEA project and described in Lutter et al. (2016). In this study, national agricultural production and associated blue water (irrigation) consumption was disaggregated to river-basin level. This was combined with water availability information to provide a blue water scarcity index (based on Hoekstra et al. 2012) associated with different crop groups. We used an allocation matrix to allocate fractions of blue water used in each basin, per crop sector, to EXIOBASE to create a nationally or regionally scaled WSI estimate which accounts for the underlying spatial landscape of crop production, water extraction and water scarcity.

The second method is conceptually more simplistic, and is not tied to sector-specific water usage or basin-level water availability. WTA ratios (freshwater withdrawal, divided by freshwater availability at national level) are drawn from FAO Aquastat, adopting the following equation from Pfister et al. (2009)

formula water scarcity copy

where WTA* is an adjusted water-to-availability ratio. For aggregate (RoW) regions in EXIOBASE, the (highly simplified) assumption is made that the WTA represents the sum of total regional freshwater extraction across countries in the region, divided by total water availability. No attempt is made to scale regional WTA values to regions of high or low agricultural productivity.

WSI values resulting from each method are multiplied by blue water use results (all for 2011) from each region/agricultural product group combination to provide water-scarcity adjusted consumption metrics. Green water use is not considered as it is not linked to irrigation, and therefore scarcity is not contextually relevant. Analysis is restricted to agricultural products as data availability constrains the first method to these sectors.

Finally, recent work has identified crop-country combinations associated with high levels of groundwater depletion (Dalin et al. 2017). Groundwater depletion presents an alternative perspective on the potential impact associated with water use (the assumption being that depletion of non-renewable resources is likely to be unsustainable). For individual crop groups, selected groundwater extraction per unit production (GWC, expressed in l/kg values for key producing regions are summarised as a complement to the WSI-adjusted results.

Results

Detailed results are presented for the first method only, reserving a short comparative discussion of results from the second method for the end of this section. A results spreadsheet accompanies this case study document.

Method one

Table 1 shows the top 20 crop-country combinations by total blue water consumption, and their adjusted rankings, for 2011. These account for 63.3% of Sweden’s total unadjusted blue water consumption, and 69.8% of WSI-adjusted consumption.

Table 1. Top 20 geographic sources of blue water embedded in Swedish consumption (for crop products), and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Crop Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
RoW Asia and Pacific Wheat 153.57 117.76 1 1
RoW Asia and Pacific Rice 103.00 58.24 2 3
RoW Middle East Fruit 70.72 62.11 3 2
China Wheat 65.74 47.93 4 5
RoW Middle East Wheat 58.81 48.49 5 4
RoW Asia and Pacific Other crops 57.39 13.00 6 20
India Other crops 46.27 24.35 7 7
United States Fodder crops 42.50 19.49 8 12
RoW Middle East Vegetables 40.46 32.96 9 6
India Sugar crops 39.64 17.21 10 14
Spain Vegetables 38.26 21.16 11 10
RoW Asia and Pacific Sugar crops 34.75 23.35 12 8
Mexico Fodder crops 34.23 16.60 13 15
RoW Middle East Nuts 29.72 18.34 14 13
RoW Asia and Pacific Fodder crops 29.58 21.70 15 9
China Rice 28.99 5.06 16 34
RoW Middle East Other cereals 28.68 20.39 17 11
Spain Fruits 25.63 14.60 18 16
India Rice 22.59 13.02 19 19
Australia Fodder crops 21.57 7.53 20 28

 

While wheat from RoW Asia and Pacific[1] remains at the top of the list after adjustment for water scarcity, there are, significant changes lower down the ranking. The crop groups that account for 10% or more of Sweden’s total embedded blue water consumption are discussed in more detail below.

Wheat

Total embedded blue water use for production of wheat associated with Swedish consumption is 313.88 Mm3 (20.5% of the total across all crops, which is 1534.18 Mm3). After adjustment for WSI, wheat’s share of total blue water use for Swedish consumption rises to 27.0% (233.24 Mm3 of 864.46 Mm3). This indicates that production of wheat for Swedish consumption is more associated with water-stressed areas than the average crop. Table 2 shows the top 10 producer regions (accounting for 98.9% of total blue water use). As can be seen, the relatively high WSIs for RoW Middle East and RoW Africa push these regions up the ranking after adjustment. Notably, Denmark-sourced wheat drops several places. Data on groundwater depletion from Dalin et al. (2017) indicates that wheat production in countries in the RoW Middle East and RoW Africa regions is also associated with high groundwater extraction per unit of production (e.g. Kuwait 21 901 l/kg, Qatar 9 389 l/kg).

Table 2. Top 10 geographic sources of embedded blue water use for production of wheat associated with Swedish consumption, totals and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
RoW Asia and Pacific 153.57 117.76 1 1
China 65.74 47.93 2 3
RoW Middle East 58.81 48.49 3 2
India 15.64 9.62 4 4
United States 6.08 1.98 5 6
RoW Africa 4.89 4.05 6 5
Russia 2.01 0.53 7 9
Turkey 1.92 0.92 8 7
Mexico 0.97 0.75 9 8
Denmark 0.89 0.09 10 15

 

Rice

Total embedded blue water use for production of rice associated with Swedish consumption is 195.59 Mm3 (12.7% of the total across all crops). WSI-adjusted total water use for rice is 101.29 Mm3 (11.7%), close to the average crop water stress profile. Table 3 shows the top 10 producer regions (accounting for 98.8% of total blue water use). Blue water use linked to production in China is relatively high (ranked second), but drops when adjusted for water scarcity. In contrast, rice from RoW Africa is associated with higher water stress after the adjustment. Within the RoW Asia and Pacific region, which ranks first with both metrics, production in Pakistan (GWC = 2 111 l/kg) is associated with a particularly high groundwater extraction rate.

Table 3. Top 10 geographic sources of embedded blue water use for production of rice associated with Swedish consumption, totals and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
RoW Asia and Pacific 103.00 58.24 1 1
China 28.99 5.06 2 5
India 22.59 13.02 3 3
RoW Africa 16.27 13.37 4 2
RoW Middle East 11.53 7.99 5 4
RoW America 3.97 0.80 6 7
United States 3.23 1.00 7 6
Italy 1.83 0.45 8 8
Russia 1.25 0.38 9 9
Turkey 0.63 0.26 10 10

 

Fruit

Total embedded blue water use for production of fruit associated with Swedish consumption is 186.32 Mm3 (12.1% of the total across all crops). Total WSI-adjusted water consumption is 120.81 Mm3 (14.0%), indicating that fruit production for Swedish consumption is slightly more associated with water-stressed areas than the average crop. Table 4 shows the top 10 fruit-producing regions for Swedish consumption (accounting for 95.4% of total embedded blue water use). Unadjusted and WSI-adjusted rankings are relatively consistent across the top 10, with Italy decreasing and China increasing. Within the GWC data, the group “fruit” is represented by three groups: citrus, dates and grapes, so an average GWC is calculated for these. This highlights the relatively high GWC in countries such as Qatar (17 163 l/kg) and Kuwait (12 292 l/kg). Spanish fruit production is associated with a relatively low level of groundwater use (6 l/kg).

Table 4. Top 10 geographic sources of embedded blue water use for production of fruit associated with Swedish consumption, totals and and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
RoW Middle East 70.72 62.10 1 1
Spain 25.63 14.60 2 2
RoW Africa 21.49 12.98 3 3
RoW America 20.75 6.46 4 5
RoW Asia and Pacific 17.78 10.26 5 4
India 11.39 6.01 6 6
United States 2.72 1.79 7 7
Italy 2.54 0.88 8 10
Greece 2.54 1.09 9 9
China 2.19 1.53 10 8

 

Other crops

Total embedded blue water use for production of “other crops” associated with Swedish consumption[2] is 161.11 Mm3 (10.5% of the total across all crops). Total WSI-adjusted blue water use is 55.67 Mm3 (6.4%) indicating that “other crops” consumption has a lower association with water-stressed areas than expected on average. Table 5 shows the top 10 producing regions for Swedish consumption (accounting for 98.8% of total embedded blue water use). Mexico and Brazil move lower in the adjusted rankings, while India, RoW Middle East and China move up. GWC data is not summarised for this classification as there is not an equivalent, comprehensive, classification in the Dalin et al. (2017) dataset.

Table 5. Top 10 geographic sources of embedded blue water use for production of “other crops” associated with Swedish consumption, totals and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
RoW Asia and Pacific 57.39 13.00 1 2
India 46.27 24.35 2 1
Mexico 12.82 2.00 3 6
Turkey 9.86 4.76 4 4
RoW Middle East 8.67 6.24 5 3
RoW Africa 6.46 0.77 6 7
United States 5.14 0.75 7 8
China 4.82 2.30 8 5
RoW America 4.70 0.53 9 9
Brazil 3.05 0.09 10 14

 

Fodder crops

Total embedded blue water use for production of fodder crops associated with Swedish consumption is 167.55 Mm3 (10.9% of the total across all crops). Total WSI-adjusted blue water use is 78.29 Mm3 (9.1% of the total), indicating a slightly lower than average water scarcity profile. Table 6 shows the top 10 fodder-producing regions for Swedish consumption (accounting for 96.7% of total embedded blue water use). RoW Asia and Pacific rises to the top of the WSI-adjusted rankings, with Sweden notably dropping. GWC data is not summarised for this classification as there is not an equivalent, comprehensive, classification in the Dalin et al (2017) dataset.

Table 6. Top 10 geographic sources of embedded blue water use for production of fodder crops associated with Swedish consumption, totals and WSI-adjusted values calculated using Lutter et al. (2016) data

Country/region Blue water use
(Mm3)
WSI-adjusted use
(Mm3)
Rank
(unadjusted)
Rank
(WSI-adjusted)
United States 42.50 19.49 1 2
Mexico 34.23 16.60 2 3
RoW Asia and Pacific 29.58 21.70 3 1
Australia 21.57 7.53 4 4
Netherlands 10.51 4.41 5 5
Sweden 7.93 0.45 6 10
South Africa 5.85 1.93 7 7
RoW America 4.01 1.09 8 8
RoW Middle East 3.42 2.94 9 6
Poland 2.43 0.40 10 12

 

Method two

This method was included in the study because accounts prepared in PRINCE are designed to be readily updateable, and it is thus appropriate to assess whether a less complex method provides an adequate proxy for more powerful methodologies.

However, comparing the results from the first method and that based on FAO Aquastat data, it is clear that there are significant differences between rankings obtained from the more sophisticated method of Lutter et al 2016. These differences are not surprisingly more pronounced when we compare results for aggregated regions, due to the fact that WTA ratios are aggregated over broad areas with no allowance made for geographic crop-production patterns. However, even when aggregate RoW regions are removed and the data are re-ranked, significant discrepancies remain, as can be seen in Table 7.

Table 7. WSI-adjusted values calculated using Lutter et al. (2016) data and FAO Aquastat data. Aggregated regions are removed and data re-ranked to show top 20 country/crop combinations (using WSI-Lutter).

Country/region Crop WSI-adjusted use
(Mm3) – Lutter et al. 2016
WSI-adjusted use
Mm3) – Aquastat
Rank –
WSI Lutter
Rank –
WSI Aquastat
China Wheat 47.93 6.65 1 6
India Other crops 24.35 15.44 2 1
Spain Vegetables 21.16 11.88 3 3
United States Fodder crops 19.49 1.97 4 13
India Sugar crops 17.21 13.22 5 2
Mexico Fodder crops 16.60 2.70 6 11
Spain Fruits 14.60 7.96 7 4
India Rice 13.02 7.54 8 5
India Wheat 9.62 5.22 9 7
Australia Fodder crops 7.53 0.32 10 20
India Oil crops 7.27 4.16 11 8
India Fruits 6.01 3.80 12 9
China Rice 5.06 2.93 13 10
Turkey Other crops 4.76 0.89 14 14
Netherlands Fodder crops 4.41 0.40 15 19
China Other cereals 4.18 0.64 16 16
China Oil crops 4.04 0.74 17 15
Spain Other cereals 3.46 2.22 18 12
United States Other cereals 3.43 0.43 19 18
United States Oil crops 3.12 0.49 20 17

 

Potential future work

The results from the first method (based on Lutter et al. 2016) could be adopted to provide an interesting scarcity-based “extension” to blue water use accounts. These indices are designed for use in MRIO modelling and provide a robust mechanism for accounting for sub-national variability in water consumption and scarcity. A caveat here is that future updates to scarcity estimates based on the methods in Lutter et al. (2016) depend on a relatively complex methodology and are thus likely to rely on third-party compilation. While the second method is simpler, and values would be relatively easy to compile, the results do not correspond well to those from the first (presumably more reliable) method. Thus, it is not recommended that the latter values are adopted.

One limitation of the Lutter et al. (2016) data is that it is only currently applicable to agricultural production, so work to extend such indicators to other sectors would be beneficial. Furthermore, recent efforts to provide linkages between sub-national trade and local water scarcity (e.g. Flach et al. 2016) offer interesting opportunities for extending this work and providing increasingly robust connections between country supply chains and regions of sub-national impact.

References

Dalin, C., Wada, Y., Kastner, T. and Puma, M. J. (2017). Groundwater depletion embedded in international food trade. Nature, 543(7647). 700–704. DOI: 10.1038/nature21403

Flach, R., Ran, Y., Godar, J., Karlberg, L. and Suavet, C. (2016). Towards more spatially explicit assessments of virtual water flows: linking local water use and scarcity to global demand of Brazilian farming commodities. Environmental Research Letters, 11(7). 075003. DOI: 10.1088/1748-9326/11/7/075003

Hoekstra, A. Y., Mekonnen, M. M., Chapagain, A. K., Mathews, R. E. and Richter, B. D. (2012). Global monthly water scarcity: blue water footprints versus blue water availability. PLoS ONE, 7(2). e32688. DOI: 10.1371/journal.pone.0032688

Lutter, S., Pfister, S., Giljum, S., Wieland, H. and Mutel, C. (2016). Spatially explicit assessment of water embodied in European trade: a product-level multi-regional input-output analysis. Global Environmental Change, 38. 171–82. DOI: 10.1016/j.gloenvcha.2016.03.001

Pfister, S., Koehler, A. and Hellweg, S. (2009). Assessing the environmental impacts of freshwater consumption in LCA. Environmental Science & Technology, 43(11). 4098–4104. DOI: 10.1021/es802423e

Vӧrӧsmarty, C. J. (2000). Global water resources: vulnerability from climate change and population growth. Science, 289(5477). 284–88. DOI: 10.1126/science.289.5477.284

 

 

[1] An explanation of the country groupings used in the PRINCE accounts can be found here.

[2] “Other crops” is all crops excluding wheat, rice, “other cereals”, roots and tubers, sugar crops, pulses, nuts, oil crops, vegetables, fruits, fibres, and fodder crops.