Climate Variability, Climate Change and Land ...

1 downloads 0 Views 793KB Size Report
Sep 7, 2006 - Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, ... land degradation in north-eastern Australian arid and semi-arid regions ...
CHAPTER 11

Climate Variability, Climate Change and Land Degradation Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

Abstract. Effective response by government and individuals to the risk of land degradation requires an understanding of regional climate variations and the impacts of climate and management on condition and productivity of land and vegetation resources. Analysis of past land degradation and climate variability provides some understanding of vulnerability to current and future climate changes and the information needs for more sustainable management. We describe experience in providing climate risk assessment information for managing for the risk of land degradation in north-eastern Australian arid and semi-arid regions used for extensive grazing. However, we note that information based on historical climate variability, which has been relied on in the past, will now also have to factor in the influence of human-induced climate change. Examples illustrate trends in climate for Australia over the past decade and the impacts on indicators of resource condition. The analysis highlights the benefits of insights into past trends and variability in rainfall and other climate variables based on extended historic databases. This understanding in turn supports more reliable regional climate projections and decision support information for governments and land managers to better manage the risk of land degradation now and in the future.

11.1 Introduction The interaction between climate, human activities and land condition is complex. The extent to which land management practices affect the condition of the land is influenced by climatic factors, and conversely, the climate restricts the range of land management practices that can be sustainably employed. These interactions, and their consequences, can lead to a deterioration of land condition which in turn has been shown to have impacts on the atmosphere and future climate. Hence, when discussing climate variations and land degradation, it is critical to be aware of the complex processes and feedbacks that occur on a range of temporal and spatial scales. Many of these processes and feedbacks are only partially understood, and so sustainable practices will inevitably be planned against a background of limited scientific understanding and uncertainties surrounding future climate changes. Nevertheless, there are good reasons to act. It is widely accepted that land use practices must minimise land degradation in order to be sustainable in the longterm. It is also widely accepted that climate variations are an important contribu-

206

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

tor to the degradation of productive and natural lands. Many of the countries most severely affected by land degradation are already economically disadvantaged and their scope to withstand any downturn in productivity is limited. However, land degradation is a global issue affecting both developed and developing nations. Global efforts to understand how climate variability has contributed to resource damage and how current and future climate change may further exacerbate the damage is critical for economic and environmental sustainability. There is a clear need for continued research into global and regional climate systems, seasonal climate forecasting, climate trends, risk assessments and communication of those risks to land managers. Decision support tools and policy development based on sound science are prerequisites for the sustainable and productive use of lands. A number of definitions of land degradation and desertification have been proposed, and the choice of definition will impact to some extent on the interpretation of the role of climate. However, the definitional question will not be addressed in this paper. We will use, as the basis for analysis, the definition of the United Nations Convention to Combat Desertification (UNCCD), i.e. land degradation is a “reduction or loss, in arid, semi-arid, and dry subhumid areas, of the biological or economic productivity and complexity of rain-fed cropland, irrigated cropland, or range, pasture, forest, and woodlands resulting from land uses or from a process or combination of processes, including processes arising from human activities and habitation patterns, such as: (i) soil erosion caused by wind and/or water; (ii) deterioration of the physical, chemical, and biological or economic properties of soil; and (iii) long-term loss of natural vegetation”. The UNCCD further describes desertification as “land degradation in the arid, semi-arid and dry subhumid areas resulting from various factors, including climatic variations and human activities”. It is the role of climatic variations that provides the focus of this paper. The socio-economic aspects of land management in dryland regions, where profitability and long-term productivity is often marginal, compound the scientific complexity of this issue. It is impossible to cover all aspects of this problem in a single paper and we do not attempt to do so. Rather, we describe one approach to analysing and addressing the relationship between climate and land degradation, derived from our experience in Australia. The United Nations Development Programme (UNDP) estimated that there are 54 million km 2 of drylands globally, representing approximately 40% of land area. About 15% of this total is in Australia, with the majority in Asia (34.4%), Africa (24.1%) and the Americas (24%). While all drylands can be considered to be at risk of degradation, the extent of degradation at any point in time is difficult to assess, again raising definitional questions (Prince 2002; Walker et al. 2002). Perhaps the scale of the problem can best be summarised in the assessment of the UNCCD that over 250 million people are directly affected by land degradation, and that one billion people, many already suffering severe poverty, in over 100 countries are at risk In the discussion that follows, we focus on the arid and semi-arid rangelands that are used for grazing domestic livestock in Australia – about 406 million ha across the continent. European settlement in Australia, beginning a little over 200 years ago, had a major impact on the landscape,. However, reliable records are available for only a shorter period (i.e. a little over 100 years). The processes of land degradation occur on a range of time scales and can be attributed to both natural

Chapter 11: Climate variability, climate change and land degradation

207

factors and human activities. Determining the causes of changes in land condition is not straightforward because natural climate variations and other, often related, natural factors such as wildfires interact with human-caused factors such as overgrazing, the introduction of animal and plant pests and inappropriate use of burning. A major study of eight well-documented episodes of degradation in Australia’s rangelands provides an understanding of the natural and human factors that contributed to those regional episodes (McKeon et al. 2004). Following a discussion of what we have learnt from analysing those historical events, we provide a description of preliminary research on how land use, land use change and climate change may exacerbate the risk of land degradation in the future. To illustrate this aspect, we present the example of a dynamical climate modelling experiment in which feedbacks cause regional land degradation to exacerbate drought conditions in central Australia. Further research aimed at a better understanding of how land degradation and climate feedbacks operate through the carbon, energy and water balances will enable dynamical climate modelling to provide more accurate climate projections in the future and so support sustainable resource management.

11.2 The risk of land degradation in Australia Agriculture remains the major land use in Australia, occupying 61.5% of the land area (7.6 million ha) with 56% used for the grazing of natural vegetation and a further 2.5% used for dryland grazing on improved pastures (SOE 2006). Australia’s rangelands are environmentally diverse landscapes and are susceptible to the impacts of rainfall variability. Rainfall is not only very low in Australia, with over 50% of the land area receiving less than 300 mm median annual rainfall, but also highly variable on timescales from intra-seasonal to multi-decadal and longer. A major driver of variations in rainfall on inter-annual timescales is the El Niño – Southern Oscillation or ENSO (e.g. Nicholls 1988), particularly in the north-east of the continent. In Australia, the combination of agricultural land use and high natural climate variability – particularly the occurrence of severe and protracted drought periods – presents a challenge for sustainable land management. This challenge has resulted in a national and regional focus on policies and research into natural resource management and climate science that provides a valuable foundation for responding to climate change. Examples include: z Australia has implemented a National Drought Policy that encourages rural land managers to manage for climate variations, but which also provides financial assistance to farmers in “Exceptional Circumstances”, i.e. in rare and severe drought events. The current definition of an exceptional circumstance is one that is rare and severe, assumed to occur on average once in 20 to 25 years, and that causes a severe downturn in farm income over a prolonged period (Botterill 2003; Day et al. 2003). z The Australian Bureau of Meteorology provides climate data and seasonal outlooks that enable farmers and graziers to proactively minimise the risk of environmental and economic damage to their properties during unfavourable

208

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

seasons. The Bureau’s information is complemented by application-driven information provided by private companies and by State governments, e.g. the Queensland Government Long Paddock website (www.Longpaddock.qld.gov. au) which provides rainfall outlooks and probabilities for pasture growth for the season ahead. z The National Agricultural Monitoring System (NAMS), coordinated by the Commonwealth government, brings together a large amount of information from national and State organisations for land managers. The NAMS information supports the adoption of risk management strategies as well as the development of submissions for Drought Exceptional Circumstances financial assistance. Responsibility for land management in Australia lies with State governments. This paper focuses on Queensland (northeast Australia) in which over 85% of the 173 M ha land area is managed for grazing domestic animals. Most of this area is semiarid to arid grasslands and woodlands. These rangelands are sensitive to the highly variable rainfall regime, with the risk of long-term degradation also linked to total grazing pressure from the combination of domestic livestock, native herbivores (particularly macropods), and feral animals including rabbits, goats and camels. Management of fire frequency, woody vegetation proliferation, weeds, and vegetation removal are additional factors in the land degradation equation. In this chapter, an integrated approach to providing climate science information to support better management of the risk of land degradation (Fig. 11.1) is described. The approach is based on: (i) maintaining a comprehensive program of land and vegetation condition monitoring using both remote sensing and field assessment; (ii) modelling rangeland systems to understand biophysical processes in historical, present and future contexts; (iii) modelling regional climate systems and providing climate projections on seasonal and longer timescales (e.g. 3 months to 50 years into the future); and (iv) engaging with government and the community to provide information on current resource condition and risks of degradation. This approach illustrates the critical aspects of managing the risk of land degradation in a variable and changing climate: z Objective assessment of land condition, and analysis of how the current conditions compare with historical conditions; z Understanding the climatic and socio-economic factors that contribute to productivity loss, degradation and recovery of land and vegetation resources; z Climate risk assessments based on seasonal and longer-term outlooks, and an understanding of the reliability of these climate projections; and z Information and education on the risk of land degradation and desertification, e.g. ‘safe’ livestock carrying capacities both now and under future climate change.

Chapter 11: Climate variability, climate change and land degradation

209

Fig. 11.1. Illustration of the multi-faceted approach used in Queensland for monitoring the risk of land degradation in extensive semi-arid and arid grazing lands and providing advice for managing the risk of land degrada-tion in a variable and changing climate

11.3 Monitoring land condition and degradation risk The objective assessment of land condition over the extensive and diverse landscapes that characterise Australia’s extensive grazing lands requires innovative monitoring techniques. In Queensland an extensive remote sensing program (Statewide Landcover and Trees Study, SLATS) uses moderate resolution (25m x 25m) Landsat imagery to detect and map woody vegetation change (clearing or regrowth) (www.nrw.qld.gov.au/SLATS). The Landsat analyses are used in combination with other products such as MODIS imagery to monitor resources on other spatial and temporal scales. Foliage projective cover and extent of bare ground are analysed to indicate rangeland condition. Bare ground (or its converse, ground cover) is sometimes used as a surrogate for land condition but time series of remotely sensed data may be difficult to interpret, even where trends in an index are detected accurately. For example, an increase in cover on grazing lands may be interpreted as improving land condition when in fact it is due to weed infestation rather than an increase in cover of edible perennial grasses. Field verification is critical to confirm the interpretation of remotelysensed imagery. Land managers need to feel confident that the remote sensing can successfully monitor the risk of land degradation, and field verification is essential in establishing acceptance of the technique. Further, experience in Queensland has shown that maintaining ongoing field data collection programmes to support the objective monitoring of land condition is essential. Transect and point measurements are used to calibrate the remote sensing programs and process models (e.g.

210

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

www.nrw.qld.gov.au/slats; Carter et al. 2000). In addition, a program of Rapid Mobile Data Collection covering the extensive area of rangelands (Hassett et al. 2006) provides a large database of quantitative, calibrated pasture attribute records as well as qualitative expert observations. Over the past 14 years, approximately half a million km were travelled as part of this program, and close to a million observations have been taken, with, on average, three attributes recorded per observation. These data provide a valuable record of field conditions and are supported by extensive photographic records. Complementary field and remote sensing programs deliver objective and reliable monitoring of the impacts of climate and management practices on the arid and semi-arid lands used for broadscale grazing. However, understanding the changes in the condition of these lands requires an understanding of biophysical processes that affect landscape conditions. In arid and semi-arid systems, available water often determines plant growth, which influences the availability of feed for domestic livestock and other herbivores, and, through ground cover, susceptibility to soil loss. A process model that reflects plant function and soil water balance is required to provide a basis for analysing how the landscape functions. Australian Grassland and Rangeland Assessment by Spatial Simulation (AussieGRASS) is a simulation model developed to monitor grass production and land cover, and to analyse the impact of climate variability on grazing lands. In the operational system, AussieGRASS runs a calibrated and validated water balance and pasture growth model on a 5 km grid across the continent, although in extensive broadscale grazing enterprises data-availability often significantly limits the accuracy of simulated values (Carter et al. 2000). AussieGRASS is based on the deterministic, point-based, daily time-step model GRASP (Figure 11.2), which simulates soil-water, pasture growth and cattle or sheep production (McKeon et al. 2000). Simulations of other indicators, including carbon stocks and grass cover, make the framework a valuable environmental calculator for rangelands. Total precipitation provides only a partial indication of drought impacts, whereas pasture response, measured as growth, biomass and cover, provides a better assessment of the impact of drought on rangelands, and a more realistic ranking of the current conditions in a historical context. Pasture growth integrates additional climatic factors such as temperature, humidity, solar radiation and the pattern of rainfall, as well as the initial condition of the ecosystem (particularly grass basal cover and soil moisture). AussieGRASS generates pasture simulations in near-real time, and the model is also linked to a seasonal climate forecast system (currently the SOI-Phase system, Stone et al. 1996) to provide outlooks for three months ahead. By taking into account livestock numbers, AussieGRASS can also assess grazing pressure, and therefore be used to indicate degradation risk and identify opportunities for improved management. Climate risk assessments provide land managers with information to support proactive decision-making and during drought the information is also used to advise the government about the risk of land degradation. Hence spatial modelling and climate risk assessment in arid and semi-arid rangelands can produce benefits at the enterprise and regional scale, as well as providing an equitable and objective assessment of pasture status in different Australian regions.

Chapter 11: Climate variability, climate change and land degradation

211

Fig. 11.2. Components of the rangelands systems, applicable to the grasslands and woodlands that characterise the extensive grazing lands of northern Australia, as simulated using the GRASP pasture growth and water balance model in the AussieGRASS spatial simulation framework

AussieGRASS is used to generate information products tailored for land managers, and for assessing the impacts of droughts. These products are made available in near-real time, and for the season ahead. However, the process-based modelling framework also makes AussieGRASS a useful generic environmental calculator. Thus AussieGRASS is used to analyse current and emerging issues of critical importance to natural resource management in Queensland such as landscape water balance, climate change impacts, carbon stocks in vegetation and soils, and impact of grazing on groundcover and sediment loss. Research is continuing to improve the accuracy of simulations and create new applications to support priority policy issues for government.

11.4 Climate information High quality climate data, as both historical information and future projections, are a critical requirement for analysing the processes of resource condition. The Australian Bureau of Meteorology (BoM) provides daily national meteorological data. The Silo project, a collaboration between the Queensland Government and

212

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

BoM, provides unbroken time-series of daily interpolated climate data from 1889 to the present. Silo was initiated in 1997 in response to the need for long time-series of daily climate data for use in biophysical modelling applications, particularly for agricultural, water and rangeland management. Silo allows spatial models such as AussieGRASS to provide analyses of not only actual land condition, but also land condition ranked relative to conditions over more than a hundred years. Expressing current land condition in relation to history is critical to support drought policy where financial assistance is based on ‘rare and extreme’ conditions. The approach is also useful for showing trends and changes over time. Time-series of historical data are also required for seasonal forecasting or climate risk assessments that are based on analogue years, such as the SOI phase system (Stone et al. 1996). Statistical modeling is used operationally to provide climate risk assessments using the AussieGRASS framework, but there is also a capability to link dynamical systems to the five km gridded surface, as described later in this chapter. Future developments in climate science will improve the accuracy of highresolution climate change scenarios, and it is may be possible to use these scenarios within AussieGRASS to simulate the impact of climate change on the risk of land degradation and on productivity (Hall et al. 1998).

11.5 Understanding past land degradation in Australia’s rangelands An analysis of eight well-documented episodes of degradation in Australia’s rangelands has provided insights into the factors that contribute to land and vegetation degradation (McKeon et al. 2004). Rainfall variability is a major driver, but land degradation is more than low moisture availability during droughts. Similarly, recovery will not occur with one good season, and managing the breaking of a drought may be as important in avoiding long-term damage to grazed landscapes as managing the onset of dry conditions. A first step in understanding the sequence of conditions that precede the damage is, therefore, to understand the pattern of climate fluctuations, as discussed in the previous section, and the pressures on the land imposed by natural and human activities.

11.5.1 Climate regime Australia’s climate reflects its location in the tropics/sub-tropics (approximately 11o S to 39o S) and relatively small land mass in relation to surrounding oceans. Table 11.1 summarises timescales of climate variations relevant to providing climate risk information for land management in north-eastern Australia. Climatic variations, particularly in rainfall, are the result of both the inherently chaotic nature of the climate system and also the effects of variations in sea surface temperatures and atmospheric circulation patterns (Fig. 11.3). The latter are to some degree predictable. Probabilistic seasonal climate outlooks in this region are based largely on

Chapter 11: Climate variability, climate change and land degradation

213

Table 11.1. Major components of climate variation relevant to land management in Australian rangelands. Climate system variation Component

Time Period

Literature Cited

Weather

Daily/Weekly

Madden-Jullian Oscillation (MJO)

Intra-seasonal (30-60 days)

Seasons

Seasonal

Quasi-biennial

2.5 years

White et al. (2003)

El Niño – Southern Oscillation (ENSO

Inter-annual (2-7 years)

Pittock (1975); Allan (1985); Nicholls(1988); Nicholls (1991)

Quasi-decadal

9-13 years

Noble & Vines (1993); White et al. (2003); Meinke et al. (2005)

Pacific Decadal Oscillation (PDO) or Inter-decadal Pacific Oscillation (IPO)

Inter-decadal

Power et al. (1999); White et al. (2003)

Multi-decadal

30-100 years

Hendy et al. (2003)

Global Warming and Greenhouse

Since late-1800s

Cai et al. (2005); Nicholls(2006)

Stratospheric Ozone Depletion

Since 1970s

Syktus (2005) ; Cai et al. (2006)

Asian Aerosols

Since 1980s

Rotstayn et al. (2006)

Land Cover Change

Since mid-1800s

Lawrence et al. (sumitted)

Milankovitch cycles or Ice Ages

Thousands of years

De Decker et al. (1988); Barrows et al. (2000)

Donald et al.(2006)

Climate Change components

an understanding of the development of the ENSO phenomenon and its impact on rainfall. The strength of the relationship between ENSO and rainfall, and consequently the reliability of ENSO-based forecasts, varies regionally and throughout the year. Farmers and graziers managing broadscale production against a background of highly variable rainfall would benefit from more reliable rainfall outlooks or climate risk assessments. However, even when the strength of signal is low, an understanding of regional climatology and the variability characteristic of the location is beneficial because it is recognised that a series of favourable years can lead to over-

214

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

Fig. 11.3. Graphic illustration of the high variability in annual rainfall in Australia’s rangelands taking the example of average values for the 12 months from April to March averaged across the western New South Wales grazing lands and expressed as a percent anomaly of the long-term mean. The patterns of inter-annual and inter-decadal variability can be related to state of ENSO and the Pacific Decadal Oscillation

expectations of livestock carrying capacity, and that contributed to past episodes of land degradation (Fig. 11.4; McKeon et al. 2004). A dominant influence on sustainable management of Australia’s rangelands is extended periods (more than three years) of above or below median rainfall (Fig. 11.3). In these sequences of above-average years, stock and other herbivore numbers have built up, e.g. in the 1880s/90s, 1920s, 1950s, 1970s, 1998-2001. If commodity prices declined rapidly, graziers tended to retain stock in the hope of an upturn in market conditions. Examples of such a decline in prices in Australia were seen in the1890s due to global depression, 1925 when wool price collapsed, 1929 due to a severe global depression, 1960s when wool and cattle price again de-

Chapter 11: Climate variability, climate change and land degradation

215

Fig. 11.4. Degradation episodes recorded in regions of Australia’s range-lands as described by McKeon et al. 2004

clined, and 1974 with a beef price collapse. High stocking rates and poor markets provided conditions more likely to result in degradation of arid or semi-arid rangelands if drought sets in. A severe drought in combination with high stock numbers can then result in loss of ground cover and resource damage. Recovery can occur relatively rapidly with a sequence of above-average rainfall years, but if there has been a loss of topsoil, depletion of seed stores, and related long-term degradation. In some cases, recovery could require decades or the resource may not return to its previous state. These analyses highlight the value of an integrated approach to assessing the risk of land degradation. Reducing stock numbers early when going into a dry period will conserve ground cover and reduce degradation. Information on current condition and seasonal outlooks provides support for more sustainable management decisions especially on suitable stocking rates. Ground cover and pasture biomass monitored using satellite imagery and field data, combined with seasonal climate forecasting systems and pasture growth modelling, provide the basis for a seasonal conditions report, climate risk assessment and drought alert. Integrating such assessment with information on stock numbers and total grazing pressure enables the degradation risk to be estimated and provide an alert system for severe

216

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

conditions. Delivery to graziers by email or website provides timely support for responsive management.

11.5.2 Climate change and land degradation Key requirements for climate risk assessment are understanding of: (1) the interaction of climate and management practices on land condition; and (2) the reliability of climate outlooks. Seasonal forecasts have been developed using statistical systems but climate change means that the past climate may not provide a reliable guide for future conditions. The threat to land and vegetation in arid and semi-arid regions, in particular, may be greater than previously experienced (IPCC 2007). Australia experienced protracted dry conditions with a strong El Niño in 2002/03 and this resulted in severe drought in much of the east and south-west of the continent (Fig. 11.5). In some of these areas, there had not been a return to average rainfall at the time of writing (February 2007). Although graziers have reduced stock numbers, the extended dry conditions have resulted in groundcover loss, tree death, and dust storms and other evidence of soil loss and land degradation (R. Hassett, Queensland Department of Natural Resources and Water, Pers. Comm.).

Fig. 11.5. Rainfall for the period April 2001 to January 2007 ranked relative to historical values since 1890 shows that large areas of Australia’s rangelands have experienced extremely dry conditions with many in the lowest percent of records

Chapter 11: Climate variability, climate change and land degradation

217

The extensive dry conditions since 2002 are only partly due to the state of the ENSO. Factors that have contributed to the decline in rainfall include very warm conditions in the central equatorial Pacific Ocean, large-scale changes in the hemispheric circulation related to stratospheric ocean depletion, and changes in land cover (Watkins 2005).

11.5.3 Future climate variations and land degradation Climate change is expected to increase vulnerability of arid and semi-arid regions to degradation. There is growing evidence that the frequency and extent of droughts has increased as a result of global warming. The fraction of land surface area experiencing drought conditions has risen from 10-15% in the early 1970s to more than 30% by early 2000 (Dai et al. 2004). There has been a general tendency towards increased precipitation in the high latitudes, particularly in the Northern Hemisphere, and decreased precipitation in semi-arid regions. A global analysis showed that abrupt changes in rainfall are more likely to occur in arid and semiarid regions, and that this susceptibility is possibly linked to strong positive feedbacks between vegetation and climate interactions (Narisma et al. 2007). It has not been possible to determine the cause of rainfall decline in eastern Australia (Nicholls 2006), but the observed trends are consistent with broader changes observed in the global sub-tropics (Dai et al. 2004; Vecchi et al. 2005). Analysis of the observed Australian climate record since 1950 shows that mean surface temperatures have increased by approximately 1oC on average, that there has been an increase in the frequency of heatwaves and a decrease in the occurrence of frosts, and that rainfall has decreased across eastern Australia and increased in the north-west (Manins et al. 2001; Smith 2004; Nicholls 2006). During the past decade there has been a strong and persistent rainfall deficit in eastern Australia and similarly the reduced rainfall conditions in the south-west corner of Australia have continued. Reductions of up to 20% in annually averaged totals are common across large regions of the continent (Fig. 11.6a). At the same time large parts of north-western Australia have experienced a significant rainfall increase of greater than 30%. The persistent changes in the pattern of rainfall over the continent have placed significant stress on ecosystems and landscapes. The decline in rainfall over this period has resulted in an even greater reduction in soil moisture and a significant reduction in environmental river flows. The reduction in available soil moisture combined with increasing surface temperatures (Fig. 11.6b) contributes to lower plant growth, loss of ground cover, thus increasing the risk of erosion. The potential flow to stream simulated by the AussieGRASS model using daily observed climate parameters shows a reduction of 40 to 60% over the continent (Fig. 11.6d). Climate risk assessment and climate forecasting skill have been shown to be important in improving decision making in the rangelands of Australia. However climate risk information for Australia has been based largely on the behaviour of ENSO. A challenge for assessing the risk of land degradation in the future is to develop a climate forecasting capability that can account for both human-induced

218

Beverley Henry, Greg McKeon, Jozef Syktus, John Carter, Ken Day, and David Rayner

Fig. 11.6. Observed and simulated anomalies in annual (a) rainfall, (b) mean surface temperature, (c) soil moisture in upper 1metre of the soil profile, (d) potential flow to stream, (e) pasture growth and (f) pasture cover, for the period 1993-2006 expressed as a percent of averages for 1970-1992. 1970-1992 is assumed to represent the long-term ‘normal’ conditions

and natural climate variations, especially on quasi-decadal to longer time-scales. It is no longer possible to assume that the next 30 years will be some random sample of historical climate variations and existing statistical systems will need to be monitored for assessing whether they retain the skill in assessing the probabilities of ENSO development and associated seasonal rainfall. New approaches are therefore needed to more accurately estimate seasonal climate risk and provide projections.

Chapter 11: Climate variability, climate change and land degradation

219

Global Climate Models (GCMs) are increasingly being used to interpret historical climate variability and to provide seasonal climate forecasts based on current sea-surface temperatures, i.e. conditions that include global warming. Interpretation of natural climate variations will benefit from improved historical datasets such as those being developed by the Atmospheric Circulation Reconstructions over the Earth (ACRE) initiative. The ACRE project involved recovery of historical instrumental surface data and using this to improve and extend the time series of digitised records. These data will then support surface observations-based reanalysis with sufficient data coverage to be valid globally to the mid-19th century (Dr Rob Allan, Hadley Centre for Climate Change, Met Office, UK, Pers. Comm., Compo et al. 2006). The reanalysis products are expected to also provide an observational basis for assessing ocean-atmosphere model integrations simulating anthropogenic effects on recent climate and allow current and future climate change impacts to be assessed against a reliable background. However, a major challenge remaining for climate scientists is to link low resolution (e.g. 2 o) climate models with historical climate data to create high resolution, e.g. (0.05o) spatial biophysical models (e.g. Syktus et al. 2003). With such models, the impacts of a range of climate change scenarios can be tested to support the adoption of appropriate adaptation strategies for future conditions. In summary, the risk of land degradation will be better managed with (1) an understanding of past trends and variability in rainfall and other climate variables; (2) plausible regional climate change projections; and (3) a basis for resource managers and government policy to more confidently change decisions in response to a likely changing climate.

Acknowledgements The contribution of many climate and rangeland scientists is gratefully acknowledged. Particular thanks go to Robert Hassett for expert advice on land condition and degradation risk across Queensland over more than a decade and Alan Peacock for preparing many of the figures. The financial support of the World Meteorological Organization (WMO), Department of Natural Resources and Water and Meat and Livestock Australia to attend the International Workshop on Climate and Land Degradation is gratefully acknowledged.

References )TTIV:2!