Assessing Resource System Vulnerability to Climate Change Methodology

In other words, resource vulnerability is a function of these factors including: climate, economic activities in the region, land users, size of resource use activities, resource use eff

Assessing Resource System Vulnerability to Climate Change: Methodology

By Yongyuan Yin1,2, Nicholas Clinton2, Bing Luo3 and Liangchung Song4

1Lead author-Adaptation and Impacts Research Group (AIRG), Environment Canada, and

Sustainable Development Research Institute (SDRI), University of British Columbia (UBC), 2029 WestMall, Vancouver, B.C Canada V6T 1Z2; Telephone: (604)822-1620; Fax: (604)822-3033; Email: yongyuan.yin@sdri.ubc.ca

2 International Institute for Earth Systems Science (ESSI), Nanjing University, Nanjing China

3 Faculty of Engineering, University of Regina, Saskatchewan, Canada

4 Meteorological Bureau of Gansu Province, Lanzhou, Gansu, China

Abstract

One challenging issue in resource planning and management is to design and apply integrated approaches to estimate resource system vulnerabilities to climate variability and change, and to identify desirable options that could be used to reduce these vulnerabilities Given the complexity and uncertainties in resource management, it will be difficult to identify those most vulnerable regions Research on assessing resource system vulnerability can provide scientific information and understanding necessary for insuring regional sustainability

Vulnerability of resources system reflects a complex set of interrelations involving biophysical, social, and economic factors on both the demand and supply sides of the resource use equation These biophysical and socio-economic factors may limit or facilitate resource supply and demand

In that sense they become determinants as to whether resources can provide various functions to meet many societal demands These functions can serve various human values, preferences, and aspirations to meet multiple demands from a variety of users such as agriculture, water, and ecosystem When a system’s propensity is to undergo impacts and lead to disruptions in its

nominal functionality as a result of climate variation or change, we assume that the system is vulnerable to climate

The current status of climate vulnerability research and vulnerability assessment show a lack of designing new methods to meet the increasing demand of policy makers The main goal of

vulnerability assessment is to develop effective methods to measure vulnerability and to assess the environmental risks in dealing with climate stresses In this study, several major factors which influence resource system vulnerability in the Heihe River region of China will be considered In other words, resource vulnerability is a function of these factors including: climate, economic activities in the region, land users, size of resource use activities, resource use efficiency, the price elasticity of supply and demand, environmental protection, policy options (economical, technical,

or policy), lifestyle associated with income increasing, and population growth

This paper focuses on methodology development for land and water system vulnerability

assessment using the Heihe River Basin as a case The paper introduces methods for the

formulation of indicators for agricultural land and water resource system vulnerability to climate variation and change Indicators are discussed in relation to their specificity, descriptive power, thresholds, and capacity for geographic allocation using ancillary or modeled data Resource system vulnerability is addressed with a description of applicable indicators, literature references, and geo-spatial data requirements for the mapping of the indicators

1 Introduction

A system’s vulnerability is related to a system’s resilience defined as the capability of the system

to maintaining its functionality in the face of a particular environmental change In this paper, the vulnerability of a system is defined as its propensity to undergo impacts or lead to disruptions in the nominal functionality of the system as a result of climate variation or change The purpose of this paper is to develop methods to assess land and water system vulnerability The paper will first highlight some major determinants of the resource vulnerability This will then serve to relate these determinants to vulnerability indicators and to focus and organize discussion on developing

appropriate methodologies for resource vulnerability assessment A vulnerability assessment

framework will be developed and applied to a case study in the Heihe River Basin in Northwestern China In particular, the paper will address the following questions:

 What are the important climatic and non-climatic exposures operating on, or expected to operate on, the land and water system of study region?

 What factors are driving the changes of the system vulnerability?

 How can vulnerability indicators be used to assess resource system vulnerabilities to present climate variations and future climate change?

 Can thresholds in climatic variables be identified which, if surpassed, would pose

substantially greater risk of harm to the land and water system or sub-systems than would

be expected if the thresholds are not surpassed?

 Does vulnerability (or adaptive capacity) to climatic exposures vary in character or degree for different sub-units of the resource system?

With barriers such as extremely fragile ecological conditions, fewer financial resources, poorer infrastructure, lower levels of education, and lesser access to technology and markets, the Heihe River region has been suffering from climate variations and will experience severer impacts of climate change on food production, water uses, and human health Moreover, the region’s

adaptive capacity is lower than in the coastal region of China The region is facing substantial and multiple stresses, including rapidly growing demands for food and water, large populations at risk

to poverty, drought, degradation of land and water quality, and other issues that may be amplified

by climate change

The time horizon considered for vulnerability assessment by AS25 project will be two: beginning with a careful assessment of the present-day climate vulnerability, which can then be used to assess resource system vulnerabilities to future climate change This paper, however, focuses on current climate vulnerability assessment

2 Major Determinants of Resource System Vulnerability

Vulnerability of resources system reflects a complex set of interrelations involving biophysical, social, and economic factors on both the demand and supply sides of the resource use equation These biophysical and socio-economic factors may limit or facilitate resource supply and demand

In that sense they become determinants as to whether resources can provide various functions to meet many societal demands These functions can serve various human values, preferences, and aspirations to meet multiple demands from a variety of users such as agriculture, water resources, and ecosystem

In this study, several major factors which influence resource system vulnerability in the Heihe River region of China will be considered In other words, resource vulnerability is a

function of these factors including: climate, economic activities in the region, land users, size of resource use activities, resource use efficiency, the price elasticity of supply and demand,

environmental protection (e.g control desertification and salinization), policy options

(economical, technical, or policy), lifestyle associated with income increasing, and population growth In addition, seasonal variations in supply and demand require the study to take account of the temporal aspect of the above factors

The biophysical functions that resource systems can provide include life-support for

biomass growth, biological diversity, and wildlife habitat; an assimilating function or resilience to absorb chemical wastes and pollutants through its biological chains and chemical cycles; and hydrological and microclimatic regulations The economic functions of resource systems are: supplying useful raw minerals and energy as inputs for economic production; food and water for human consumption In addition, resource systems provide a stream of social functions that are essential for supporting human welfare such as housing, employment, defence, recreation, health, cultural, scientific, educational, and aesthetic services (d'Arge, 1972)

2.1 Biophysical Determinants of Land Resource System Vulnerability

In resource systems, natural ecosystems are transformed into hybrid agro-ecosystems or water use system for the purpose of food, fiber, and other production These hybrid systems are directly dependent on biophysical factors and essential ecological functions for sustainability (Conway, 1985) The major biophysical determinants that are related to the land and water use are listed in Figure 1 (Biophysical System) Climate variables (temperature and precipitation), the slope of land,

as well as soil type and fertility, are some critical biophysical determinants relating to the supply of land for agriculture and forestry The quantity, quality, and distribution of resources, and the way they are utilized, are other essential factors for resource use decision making

The availability of land and water resources for meeting increasing demand in each

particular area is limited The existence of possible absolute biophysical constraints on resource use activities implies some limits to exploit the resource base Therefore, the alleviation of the threat of resource scarcity requires that resources be used within their biophysical limits or capability (Page, 1977; Daly, 1984)

There is increased concern about the effects of climate change arising from economic activity on the availability and suitability of resource systems for agriculture, forestry, and wildlife (IPCC, 2001; Parry et al., 1987) Climate change by changing the biophysical determinants of resource use could most directly affect various functions of resources Changes in these biophysical determinants would likely result in negative impacts on productive, hydrological, and other

functions of the land resource system The implication is that climate change will affect the supply

of resources with respect to their availability, suitability, and distribution

2.2 Economic Determinants of Resource Use

Economic determinants provide another set of opportunities or limitations to the resource use (Figure 1, Economic System) Advanced technology and managerial skill have raised land

productivity In turn, this may have eased resource system vulnerabilities If future technology and managerial skill can continue to improve the productivity of resource systems at a rate faster than the growth rate in demand, no additional resources will be needed for production (Pierce, 1990)

Population growth and urbanization will contribute to the growth in demand for resources The earliest popular description of the concept of carrying capacity as related to human population was by Thomas Malthus and may be referred to as the limit of a given resource system to support the demand and consumption levels associated with a human population This concept has also been applied to establish economic, social, and behavioral thresholds beyond which the environmental quality will deteriorate and user enjoyment will decline (Mitchell, 1989)

With high standards of living, people in developed countries demand low density housing which often translates into sprawling urban development and costly services In addition, modern societies desire more open space for recreation and public parks Many of these developments occur

in productive farmlands, forestry lands, and areas which are perceived to be of natural, historical, cultural, scenic, or scientific importance Economic development and urbanization in both

developed and developing countries are often accompanied by increasing stress on the resource system and cause significant adverse effects on the ecosystem (Hufschmidt, 1983) Many economic development activities do not pay sufficient attention to resource depletion and environmental

deterioration Resource degradation, such as erosion, salinization, desertification, and water

pollution has caused a decline in crop yields and an increase in production and environmental costs

2.3 Social Determinants of Land Use

People with different cultural and historical backgrounds perceive and value resources in different

ways Moreover, their perceptions and valuations with respect to land and water resources have

changed over time (Rees, 1985; Tisdell, 1989) Different perceptions and valuations often lead to

alternative land and water uses For example, the environmental movement has influenced resource

use patterns through an increase set aside of wilderness areas Other social factors affecting resource

use include policy, development programs, government regulation, inter and intra-generational

equality issues (Figure 1, Social System)

Characteristics: Variables: Characteristics Variables Characteristics Variables:

Biogeochemical and

hydrological

cycling,

habitat,

bio-diversity,

resilience,

production,

amenity

Climate, soil, resource quantity and quality, location or

distribution

Net Return, stability of income, food, fibre and water supply timber, mineral,

and energy supply

Relative location, resource use pattern, population, technology, managerial skill

Housing, employment, defence, public park, education.

Land tenure, policy, regulation and program, equality, living standard

Figure 1 Determinants (Variables) of Resource System Vulnerabilities

3 Assessing Resource System Vulnerability

The main goal of vulnerability assessment is to develop effective methods to measure

vulnerability and to assess the environmental risks in dealing with climate stresses In this respect,

some methods are presented below for assessing agricultural land and water system vulnerability

3.1 Resource vulnerability indicators

In resource vulnerability assessment, indicators are used as decision criteria or standards by which

the degree or the group of resource vulnerability class can be identified Efforts to assess resource

vulnerability must first to identify and specify multiple indicators Some operationally useful key

indicators in vulnerability and adaptive capacity assessment are listed in Table 1

Table 1 Potential determinants (climate and other variables with the forcing) and resource

vulnerability indicators in Heihe River region

Climate and other related

determinants (forcing)

Related system attributes and options Resource vulnerability indicators Rainfall - variability

Drought

Temperature - max

Soil moisture

Water flows, storage volumes, and quality

Palmer drought severity index Evaporation

Food sufficiency ratio Farm income

Water scarcity (withdraw ratio) Drought hazards

Resource Vulnerability Indicators: food sufficiency, water supply, income, arable land

protection, soil erosion control, ecological footprint, wetland protection, water conservation

Resource Vulnerabilities

Temperature - min

Wind

Cold snap

Heat stress days

Accumulated degree days

Cropping area

Population growth

Economic growth

Technology

Consumption

Urban expansion

Resource management

Government policies

Soil moisture Irrigation Land conversion Land use plan Adaptive capacity Adaptation options and policies

Groundwater stress Hydro power Arable land loss Salinity

Soil erosion Grassland deterioration Water quality

Wetland area Water use conflicts

CO2 and CH4 emission

It is obvious that economic return is one of the most important indicators for measuring

vulnerability and adaptive capacity Developed countries possess high adaptive capacity

Improvements in economic return will enhance adaptive capacity

In China, providing enough food for the 1.3 billion population is always a big challenge There has been an increasing concern about China’s food supply and its ability to feed itself The provision of adequate food on a continual basis is a major indicator of regional sustainability The agricultural production indicator reflects the ability of the land base to maintain in perpetuity a given flow of goods and services Agricultural production can also be considered as a security indicator to achieve higher levels of self-sufficiency and/or it may be used to represent a

vulnerability indicator to check whether the resource base can provide enough food supply

Many of the industrial and housing developments occur in productive farmlands, forestry lands, and wildness areas How to slow down the conversion of farmland to urban and industrial uses is critical for regional sustainability in China Thus, a further indicator to protect and

conserve arable land reflects this concern

It is now generally realized that an environmental concern should be incorporated in decision making in an effort to achieve sustainable development (WCED, 1987) There are a large number of parameters that can be used as indicators of ecological vulnerability For example, environmental concern may mean protection natural resources, or it may mean minimizing the concentration of atmospheric carbon dioxide at a global scale In Western China, the

environmental concern is reflected in the indicators of soil erosion, desertification, greenhouse gas (CO2, CH4) emission, and sand storm

There is an increasing concern about the implications of climate change for water

management (Gleick, 1990) There is an increasing concern about water use conflicts in semi-arid region of Western China Dealing with potential water use conflicts with changing climate is therefore considered as an important indicator The fight over access to water resources in the Heihe Basin has led to disputes, confrontation, and many cases of violent clashes The growing water use conflicts have posed a big challenge for Chinese government agencies to implement some effective water allocation policies Global warming may change average and extreme high and low river flow Changing water supply induced by climate warming may increase water use conflicts in the region

3.2 Identifying Critical Thresholds for Indicators

The critical thresholds (CT) for indicators will be set to compare with indicator values of different areas to identify there vulnerability levels against these indicators If indicator values do not exceed the threshold level, we assume that the system will have relatively benign experience and beyond which the system will feel significant stress under climate variation and/or change

For example, drought hazards are based on rainfall amount, or aridity index, and if

conditions remain below the threshold levels for a sustained period, drought hazards are declared The threshold levels (drought index) can be used as criteria in measuring the frequency of drought hazards over time The thresholds can also be used to measure the level of damages that droughts may cause Another example is annual water withdraw ratio indicator WMO suggests that the withdraw ratio exceeds 20% and 40% of annual water availability be considered as medium and high water stress respectively In Northern China, however, the threshold of this indicator is much higher, a high stress level at 60%

In vulnerability and adaptive capacity measurement, many of the indicators can be expressed

in numerical terms, particularly for those climate and physical variables It is also recognized, however, many indicators cannot be quantified, and many of the threshold levels can only be qualitatively described For instance, with respect to the CT level of the soil erosion indicator, the soil loss tolerance value, or T-value, can be used as a target level of the soil loss indicator Soil loss tolerance is a useful concept in the relationship between erosion and productivity The degree to which a unit quantity of soil loss reduces yield is dependent on a range of soil characteristics, which may be summarised as "soil loss tolerance" Soils with a concentrated distribution of nutrients in the topsoil, and shallow rooting depths, are usually sensitive in yield loss to soil erosion, and thus will have a low T-value This denotes a soil with low tolerance to erosion Soil with good structure, and deeply weathered with good nutrient reserves, will be less sensitive to erosion, and thus have a higher T-value (FAO, 1984) The T-value expresses a "tolerable" soil loss limit in order to retain productivity of the soil affected by erosion

Yohe and Tol (2002) suggest that the relationships between adaptive capacity and its determinants are difficult to quantify Functional representations of these relations are only useful when they can offer insights to the complexity Many of the vulnerability and/or adaptive capacity indicators cannot be quantified, and many of the functions can only be qualitatively described With this respect, stakeholders, policy makers and analysts jointly identify CT levels is commonly used CT levels can be specified using a range of simple to very complicated methods As

mentioned earlier, CT level for drought indicator can be determined by the amount of rainfall required in a specific region It also can be set using complex way such as the accumulated deficit

in irrigation allocations over a number of seasons (Jones and Page, 2001)

3.3 Measure Vulnerability

Resource system vulnerability is closely linked to environmental risk which can be expressed in the following simple formulas:

Environmental Risk (ER) = exposure (e) frequency (probability)  consequence

Consequence = f{intensity, sensitivity (s), adaptive capacity (a)}

Where: The frequency or probability of an environmental stress can be expressed as the likelihood

of a specific hazard (e.g climate extreme) The consequence is the damage or adverse impacts of the environmental stress It is the function of intensity of the stress, and the sensitivity and

adaptive capacity of the exposure system

The resource system vulnerability can therefore be expressed by the relationship:

Vulnerability = P(s)*P(e)*[1-P(a)]

Where: P is probability With zero probability of any one of these factors, it can be suggested that the system is not vulnerable

In many cases, however, we cannot obtain quantitative data of the probability distribution

function for these factors Linguistic representations could be used, such as very frequent,

reasonably probable, unlikely and extremely unlikely, to reflect the probability parameters It is

also often to assign numbers (e.g extremely unlikely could be assigned to a 1 in 1000 year event).

In fact, the assumption that vulnerability can somehow be expressed in terms of various

combinations of these three attributes is a necessary simplification for investigation of

vulnerability in a variety of forms

Current vulnerability can be expressed as a statistical measure of the extent or duration of

a resource system failure under climate stresses, should a failure occur The extent of a system failure is the amount an observed statistic value exceeds or falls short of the critical threshold For example, agricultural system vulnerability can be measured to show a system's failure to meet an operational goal, such as food supply for a region, or continually generating income above a minimum level for farmers

Water system vulnerability can be measured with a system's success or failure to meet demand for a certain amount of water for a municipality, or to continually release water above a minimum flow rate from a reservoir Water system vulnerability can also be measured as the average deficit occurring during failures to meet a target, as well as the severity of failures If we use river flow, F, as an indicator to measure vulnerability, the water system vulnerability can be calculated by:

EVf = Max [0, LFt-Ft, Ft-UFt]

Where: EVf is water system’s maximum-extent vulnerability based on river flow indicator; LFt and UFt are the lower and upper critical thresholds of the coping range respectively; and Ft is the observed river flow data

If the observed data (system performance indicator values over time) are within the upper and lower thresholds (within the coping range), we will then assume that the range of values are satisfactory, acceptable or un-vulnerable Statistics or observed data above the upper threshold or below the lower limit are considered as unsatisfactory or vulnerable It should be noted that these copying ranges may change over time

Vulnerability can also be measured as the maximum duration of failure The Maximum Duration-Vulnerability can be calculated by:

Maximum Duration-Vulnerability (p) of DVf = Maximum duration (number of time periods) of a continuous series of failure events for indicator f, occurring with probability p or that may be exceeded with probability 1-p

Other methods for calculating vulnerabilities of some key indicators are listed below for

illustration purpose

 Average annual water withdrawal ratio can be used to identify those sub-units which are under water stress

 Reservoir system vulnerability defined as the magnitude of a water supply failure as a fraction of annual yield can be computed by: Vrf = 0.452 * (S/Y)1.27 (Where: Vrf is reservoir vulnerability, S is the reservoir storage capacity, and Y is annual reservoir yield)

 The FAO CROPWAT model can be used to estimate some critical values of crop growth and water requirements The computation of indicators of crop stress or yield index can be achieved by using (Allen et al 1999):

Yield Index = ETc-stressed/ETc-max

 With respect to land system vulnerability, the soil erosion rate can be used as another

indicator Soil loss rate can be calculated by Universal Soil Loss Equation (USLE) or wind erosion model The soil loss tolerance value can be used as the threshold level Soil loss tolerance is a concept in the relationship between erosion and productivity

Once the vulnerability measures of the time-series values are defined, they can be applied to project indicator values over years into the future This can estimate resource system vulnerability under climate change scenarios If over time the measures of vulnerability are increasing, the resource system is getting more vulnerable When we use a set of vulnerability indicators to measure a system’s vulnerability, it is possible that vulnerability results of some indicators will be improving while others will be worsening

3.4 Vulnerability Classification by the Fuzzy Set Model

A fuzzy pattern recognition model can be applied to calculating the aggregated ratings of

vulnerability ranks for all the land units in the region In the vulnerability fuzzy classification, each indicator is considered to be a fuzzy criterion, and the membership function is expressed in parametric form The advantage of fuzzy classification is its strength in relating many criteria simultaneously Through pattern recognition modelling, the general risk level of a land unit is identified by analyzing a generated fuzzy vector The vector also contains the membership

degrees of other risk levels which provide further information on the exposure risks

3.5 Mapping Vulnerability

Mapping the spatial distribution of system vulnerability is part of the study to facilitate policy makers identifying the most vulnerable sub-units To geographically distribute the indicators, several spatial scales have been considered ranging from square kilometre, county level, sub-basin, to the whole basin based on data availability and other logistical reasons Thus data that represent various system vulnerabilities may be mapped at different scales

4 Vulnerability Assessments in the Heihe River Basin

The Heihe River Basin case study of vulnerability assessments is presented below for illustration purpose It should be noted that results presented in this chapter are mainly from current

vulnerability assessment This section provides information on the geographical distribution of current climate vulnerability levels of the region Thus, the following applications show how

methods described in section 3 can be employed in vulnerability assessment

In this section, several vulnerability indicators listed in Table 1 are selected to measure

resource vulnerability under current climate condition Geographic information system (GIS) or mapping tool is employed to identify the spatial distribution of system vulnerability in the region Maps, tables, and figures provide visual displays of resource system vulnerability, which can facilitate policy makers to identify the most vulnerable sub-units It should be indicated that there have been few practical applications of such approach yet particularly in climate vulnerability research In this sense, the methodology developed by this study can provide an introduction to a useful computer technique for climate vulnerability assessment

Results of current vulnerability assessment will establish a baseline set of measurements and observations that can be used to measure progress toward reducing vulnerability to future climate change Once these vulnerability measures are identified for various vulnerability

indicators, they can be applied to project potential vulnerabilities of the resource systems to future climate change scenarios Thus, the research on present vulnerabilities and adaptive capacities of resource systems will provide insights into potential impacts and vulnerabilities associated with future climate change While results shown in the case study are mainly based on current

conditions, the research team of the AS25 project is now applying methods to investigate climate change vulnerabilities in the region

4.1 The Heihe Region

The Basin is located in a region with the latitude of 35.4-43.5°N and the longitude of 96.45-102.8°E Figure 2 is a map of the study region The study area is the second largest inland river

basin in the arid region of Northwestern China The Basin includes parts of two provinces

(Qinghai and Gansu) and Inner Mongolia Autonomous Region With an area of 128,000 square kilometres, the region is composed of diverse ecosystems including mountain, oasis, forest, grassland, and desert Heihe River flows from a headwater on Qilian Mountain area to an alluvial plain with oasis agriculture, and then inters deserts in Inner Mongolia, representing the upper, middle and lower reaches of the Basin respectively

Figure 2 Map of the Heihe River Basin

4.2 Measuring and Mapping Vulnerability in the Heihe Region

The Palmer Drought Severity Index

The Palmer drought severity index (PDSI) was introduced by Palmer (1965) for measurement of meteorological drought It has been widely used in different regions of the world to study severity

of drought hazards (Briffa et al., 1994; Kothavala, 1999; Ntale and Gan, 2003) Because PDSI can simulate monthly soil moisture content, it is thus suitable to compare the severity of drought events among regions with different climate zones and seasons (Makra et al., 2002)

The computation of the PDSI begins with a climatic water balance using historic records of monthly precipitation and temperature Soil moisture storage is considered by dividing the soil profile into two layers The indicator operates on a monthly time series of precipitation and temperature to produce a single numerical value between +4 and -4 that represents the severity of wetness or aridity for a particular month Any PDSI values above +4.00 or below -4.00 fall into the "extreme" category of wet spell or drought

Figure 3 illustrates the trend in growing season PDSI for lower reach of the study basin It shows that the lower reach area has been dryer in the past decade This trend would continue under the changing climate Figure 4 illustrates the trend in growing season PDSI for middle reach-lower part of the study basin It also shows that this area has a trend of becoming dryer in the past decade Figure 5 illustrates the trend in growing season PDSI for middle reach-upper part

of the study basin The drought trend for this area in the past decade is not very obvious since this area is much close to the high mountains and annual average precipitation is relatively higher Figure 6 illustrates the trend in growing season PDSI for upper reach of the study basin The drought trend in this area has little changes in the past decades due to higher annual average precipitation This has particular meaning for the whole basin since most of the water resources in the study basin are from the upper reach of the basin

-2

0

2

4

6

8

10

19611964196719701973197619791982198519881991199419972000

Year Growing season PDSI

growing season PDSI Poly (growing season PDSI)

Figure 3 Trend in growing season PDSI for lower reach of the study basin

Figure 4 Trend in growing season PDSI for middle reach-lower part of the study basin

-6

-4

-2

0

2

4

6

1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000

Year Poly (growing season PDSI) growing season PDSI