October 2008

Managing water resources under changing climate: experiences from California

By Francis Ilwhan Chung

This article is based on his lecture given to the 60th session of the WMO Executive Council on 26 June 2008, Geneva, Switzerland


Only a small fraction, 2.5 per cent, of all water on Earth is freshwater. Almost 70 per cent of that freshwater is frozen in the icecaps of the Antarctic and Greenland; most of the remainder is present as soil moisture or lies in deep groundwater inaccessible for human consumption. Far less than 1 per cent—or, to be a bit more precise, about 0.007 per cent of all water on Earth is accessible for human use. This is the water that we find in lakes, rivers, reservoirs or in accessible groundwater basins. Defining water scarcity as having less than 1 700 cubic metres per capita per year of fresh water, the map below shows regions of current vulnerabilities of freshwater resources and their management.


Examples of current vulnerabilities of freshwater resources and their management; in the background, a water stress map based on Alcamo et al. (2003a). (Figure 3.2, Chapter 3, IPCC Fourth Assessment Report, Working Group 2 Report: Impacts, Adaptation and Vulnerability
Click for enlargement


Similar information compiled by the Water Resources Institute shows that low-income nations are especially vulnerable to water scarcity. Population growth will further exacerbate water scarcity.

There is a saying that in the American West “Whiskey is for drinking and water is for fighting over.” Water is called “liquid gold” in California. The current population of California is about 38 million people and is projected to be close to 60 million by 2050. California’s Gross Domestic Product is the largest of the USA. And if you count California as a country, California’s GDP is number seven in the world. Precipitation in California is largely concentrated in the northern part of the state, whereas the southern part is largely dry, less than 5 cm of annual rainfall in the south-eastern part. The major population centres—San Francisco, Los Angeles and San Diego—are located along the beautiful coast of California, away from the water source in the high mountain range. Productive agricultural lands are located in the central valley—again away from the water source, requiring man-made irrigation. California has cool, wet winters and dry, hot summers—a Mediterranean climate. Precipitation takes place during the winter, whereas the agricultural demand peaks during the summer.

California has a long history of water development going back 150 years. Early settlers relied on local runoff and local groundwater supplies. They soon found out that local supplies were not enough to meet the need of growing urban demands. The first major water movement was through the Los Angeles aqueduct, bringing snowmelt water from the eastern mountain range to the city of Los Angeles. In the same year, we saw another westward water movement, again from the Sierra Mountain Range, to the city of San Francisco. The Mokelumne Aqueduct and Colorado Aqueduct were soon added to continue moving water to the western parts of the state. In spite of the construction and operation of these westward aqueducts, the state quickly found that there was not enough water. This time, the state and federal governments became involved and started moving water from the north to the south. All of these efforts were made long before climate change was recognized as a major challenge.

The IPCC Fourth Assessment Report, released last year, projects tht global temperature may rise by 1.8 to 4°C by the year 2100, and that sea level may rise by 0.6 m. The change in precipitation is less certain, but will likely increase at higher latitudes and near the Equator. Extreme events such as floods are expected to be more frequent.

The major impact of rising temperature on water resources is two-fold. First, it is on precipitation. The rising temperature will cause the form of some precipitation to shift from snow to rain. This is especially important for areas like California that depend on snowpack for water supply. The timing and intensity of precipitation may also change. The other major challenge of rising temperature on water resources would be sea-level rise. Rising sea level may impact coastlines, bays and estuaries. Various hydraulic structures and water facilities may need to be relocated or measures to protect them need to be developed and implemented.

The figure below from the same IPCC report, shows the projected reduction in annual runoff by 2041-2060 relative to 1900-1970, in per cent, under the SRES A1B emissions scenario and based on an ensemble of 12 climate models. Models agree on the direction of runoff change in the high latitudes of North America and Eurasia, with increases of 10 to 40 per cent. Decreases in runoff of 10-30 per cent are observed in the Mediterranean, southern Africa, and western USA/northern Mexico.

world map  

Change in annual runoff by 2041-2060 relative to 1900-1970, in percent, under the SRES A1B emissions scenario and based on an ensemble of 12 climate models. Reprinted by permission from Macmillan Publishers Ltd. [Nature] (Milly et al., 2005), copyright 2005. (Figure 3.4, Chapter 3, IPCC Fourth Assessment Report, Working Group 2 Report: Impacts, Adaptation and Vulnerability)

In California, we have identified initially six major areas that will be impacted by changing climate. The first area would be the water supply, the result of changing hydrology. Warmer atmospheric temperature may cause higher water consumption by crops, natural vegetation, and humans. The rising air temperature will trigger higher water temperatures. Warmer water means more challenges in managing the water quality. That, in turn, will have impacts on fish species and the ecosystem at large. The operation of dams, reservoirs, pumping plants and hydropower plants will be impacted by the changing climate. And last, but not least, would be the threat of flood events. Both the intensity and frequency of peak flow events are expected to increase.

Agriculture is responsible for 87 per cent of the total water used globally. A recent landmark study reviewing over 1 000 papers finds that climate change is bad news for US agriculture. Published in May 2008, the Synthesis and Assessment Product 4.3: The Effects of Climate Change on Agriculture, Land Resources, Water Resources, and Biodiversity in the United States is the most extensive examination of the impacts of climate change on important US ecosystems undertaken to date. Some of the study findings forewarn bad news for agriculture.

  • Increased temperature will increase the risk of crop failures
  • Higher temperature will negatively affect livestock
  • Size and frequency of forest fires will increase
  • Insect outbreaks and tree mortality will increase
  • Increased drought conditions in the west and south-west USA
  • Weeds grow more rapidly under elevated atmospheric carbon dioxide

In 2005, during the UN World Environment Day Conference in San Francisco, the current Governor of California, Arnold Schwarzenegger, made an announcement, effectively ending the debate in California on whether climate change is real or not. His subsequent executive order directed the California state government to move ahead to develop a plan to deal with climate change.

The Governor set three numeric goals for greenhouse-gas emission reductions. The first target is, by 2010, to reduce the emissions to the level of year 2000. This means the reduction of 59 million metric tonnes of greenhouse gas, or 11 per cent below projected “business as usual.” The second set of goals is to reduce to the 1990 level by the year 2020 or by 145 million tonnes, 25 per cent below “business as usual” projection. By 2050, the reduction goal aims for 80 per cent below the 1990 level of emissions.

In order to implement the Governor’s executive order, the Climate Action Team was formed within the state government. Its main task is to develop strategies and implement actions on two fronts—mitigation and adaptation. Initially, five major sectors of adaptation were specified and water was at the forefront of the adaptation sectors.

In response to this executive order, the California State Department of Water Resources produced this report on managing water. The approach taken can be summed up in three major steps. The first step was to show the changing trend in the recorded hydrology before you even get to any mathematical climate models or any future projections. The second step was to evaluate different models before using them to project the future climate. The so-called, “re-analysis period”, which spanned 1950 to 2000, was used for model evaluation purposes. And the last step, of course, was to use the projections coming out of these different climate models to evaluate water resources system response.

So the first effort is to examine historical records. The figure below shows temperature trend in various parts of the California grouped by population size. It shows warming across the board with the most warming in urban areas. While there is, indeed, a “heat island” effect: larger urban areas tend to be warmer, but urban areas of all sizes show increase in temperature.


Warming in urban areas: the top, red line represents cities with more than 1 million people; the middle, green line cities with 100 000-1 000 000 people; and the bottom, blue line, cities with fewer than 100 000 people.

Statewide precipitation, however, does not show any definite trend, either up or down. A closer examination of the chart may show that the variations in the latter half of the 20th century tend to expand.

Another way to display this expanded variation can be shown by more frequent events of extremes. For two 30-year blocks—1921-1950, and 1970-2000—the numbers of wet years and critically dry years have increased.

The figure below shows the portion of flows during April-July as the percentage of the whole year runoff. The spring portion of the annual flow is in decline. We attribute this decline in spring flow to the reduction in the snow form of precipitation and the early melting of snow pack due to rising air temperatures.

graph   The percentage of Sacramento River annual runoff occurring in April-July trend indicates a reduction in spring runoff

Moving forward into future projections, the California temperature and the precipitation projections by different climate models with different emissions scenarios are shown up to 2100. Projected precipitation is all over the map showing no definite trend, whereas temperature on the upper chart shows a rather strong upward trend.

Projections on the future snow pack by the month of April show dramatic decreases. By 2030, this projection estimates that 95 per cent of snow water equivalent will be left, by 2060 what’s left will be 64 per cent and, by 2090, the projection is 52 per cent reduction in snow water equivalent—more than half the historic snow water equivalent will disappear.

The figure below is a view of the San Francisco Bay Delta System. The city of San Francisco is located at the tip of this peninsula, the cities of Berkeley and Oakland are located on the east side of the San Francisco Bay, the sea-level gauge is located right in the northern tip of this peninsula.

san francisco

The gauge shows a rising trend for the last 100 years of about 25 cm. That is a major concern. First we are concerned about coastal erosion and the resulting floods in low-lying areas. For example, this is one of the locations affected by the ocean tide. A rise of some 30 cm in sea level could convert a 100-year flood event into a 10-year event. According to a recent study, with a 1-metre rise in sea level, about 300 square kilometres will be newly at risk of monthly inundation.

The impacts of sea-level rise on a delta region will be depend on the shape, slope, soil type and land uses of an estuary. The Ganges Delta has a quite different geographical shape from the San Francisco Bay Delta. Rising sea levels can cause sea-water intrusion into freshwater sources in rivers or groundwater storage, floods, levee failure, land inundation or habitat losses.

Let us now focus on how future climate projections by various global climate models (GCMs) were used in California in evaluating the water resource system response. GCMs with various emission scenarios are first downscaled. The downscaled precipitations and temperature are then processed through hydrological models to generate inflows to operational models. Operational models are used to evaluate the system sensitivities and possible future facilities operations. At this stage of analysis, the systems model evaluates the reservoir operation, water temperature, river flows and hydropower generation. In California, the aforementioned bay-delta system, because of its importance, is closely examined to assess the impacts on ecosystems, water supply, water quality and the consequences of sea-level rise.

In the 2006 report, two global circulation models, together with two emissions scenarios were chosen four scenarios. Changes in inflows at various reservoirs are shown in the figure below. Seasonal shifting of runoff is apparent, while the annual volume shows decline in three out of four scenarios.

map   2050 Change in reservoir inflows

This chart shows one sample output describing the water supply aspect of four climate scenarios.

What’s up for 2008 updates?

We are expanding from three to six GCMs, but we shall retain to the same two emissions scenarios, A2 and B1.

We are adding one more downscaling technique to the one that we used in 2006, so we will have an opportunity to gauge the impacts of different downscaling techniques.

We are taking a step further on sea-level rise investigation as we are concerned about salinity increases in the San Francisco Bay delta estuary, increased flood potential, levee stability and impacts on the ecosystem. Here we see three different sea-level projections, the IPCC Third and Fourth Assessment Reports and the recent work done by Dr Rahmstorf. Again, they show a wide range of sea level projections. Dr Rahmstorf was kind enough to provide his computer code to my staff and we were able to reproduce his results published in Science magazine.

He related global sea-level rise to global mean surface temperature and discovered that 3.4 mm per year per degree Celsius rise is justified. He is projecting a rise of 0.5 m to 1 m by the year 2100.

The figure below shows how we in California translate the sea-level rise in operational planning models. With the help of the artificial neural network technique, we translate the effects of sea-level rise to the salinity increases at various locations in the estuary. This information is then fed into an operational model like the California Simulation Model, CalSim.

illustration   Artificial neural networks for sea level rise

I would like to share with four thoughts with readers:

  • Think global, act local, share global
  • Move from scientific knowledge to public policy implementation
  • Coordinate and integrate adaptation actions for multiple sectors
  • Distinguish, but integrate mitigation and adaptation actions

Climate change is a global process. Therefore, it is crucial to think globally in our collective wisdom. The actions we take, however, are at the local or regional scale. Uncertainties are abounding: different climate projections by various global circulation models, different downscaling techniques, uncertain future populations, unknown future societal value, unpredictable future economic and financial conditions or unknown future technology development. These unknowns together all pose a serious challenge in planning and implementing adaptive actions. Regional variability of future climate impacts on water resources, as well as differing regional capabilities in combating the future changes, will necessitate a worldwide coordination. I call on the climate science community to continue its efforts in developing probabilistic future hydrological projections and thereby minimize the future uncertainties. I do realize, however, that, in spite of the very best efforts, uncertainties will not be completely eliminated. Therefore, planning methods under uncertainties with non-stationary future hydrologic events must be continuously developed and refined. The knowledge and experience that we gain locally must be shared globally to help us better prepare globally. Assistance to developing countries with data, methods and other resources will be a must for this worldwide endeavour to be successful. WMO can and will play a vital role in providing technical assistance to developing countries and sharing successful local experiences worldwide.

Let’s move from science to real life implementation. I find it useful to reflect on the so-called Keeling Curve showing the variation in concentration of atmospheric carbon dioxide at Mauna Loa, Hawaii, since 1958. Through Keeling's foresight and undying perseverance, we now have his famous and insightful curve. A vast volume of scientific research and effective public campaign like the work of IPCC and Nobel Laureate Al Gore has convinced many in the public. In California, a recent public poll suggests that 65 per cent of residents believe that climate change is real and appropriate actions must be taken. A clear and firm political leadership like that of Arnold Schwarzenegger has been instrumental in moving beyond studies and discussions. In California, we believe the time is ripe to move from scientific knowledge to public policy implementation. A number of knowledge gaps still need to be filled in for this transition to take place. Setting the planning horizons, making decisions under uncertainties, conducting economic analyses for proposed adaptation actions, staging projects over time as we learn more with greater certainty, raising funds for actions and establishing collective public actions, are some of the key tasks ahead of us.

For adaptation actions, we can number several key fronts—water, agriculture, public health, coastline, forestry and ecosystems. I briefly mentioned earlier the connection between water and agriculture. Our adaptation strategies on all of these fronts must be coordinated and integrated so that the collective actions will be effective in results and cost-efficient.

Finally, I suggest the time has come to distinguish adaptation actions from mitigation actions. At the same time, however, it is essential to integrate these two types of response to changing climate. For instance, in the case of sea-level rise, the time focus for certain adaptation actions can be 40-50 years, whereas 100 years or longer for certain mitigation actions will not be beyond the norm. Adaptation actions must be devised and implemented in view of the mitigation goals. For instance, carbon footprints for certain water-conservation measures should be considered and quantified so that their advantages and disadvantages towards achieving mitigation goals and objectives can be evaluated. Likewise, mitigation actions should consider future adaptive actions. For this integration to take place, it is important that a comprehensive action plan is developed in advance and managed adaptively throughout the implementation stage.

Managing water in California was a major challenge even before awareness about changing climate. Climate change is adding another layer of complexity and uncertainty in managing water in California and worldwide. I am hopeful, however, that this new challenge will bring together people with creative ideas in their collective wisdom and, hopefully, we will all become efficient and effective in managing our precious water resources in the future.

California Department of Water Resources, Sacramento, California, USA

back to top


Condensed version/
version condensée
cover   cover
English   Français

Regular features
In the news
Recently issued
Upcoming events
Recent events
High-impact weather events
50 years ago

MeteoWorld archive


Contact: MeteoWorld Editor - WMO ©2008 Geneva, Switzerland