In June 2009, a group of experts in climate science, crop modeling, and crop development gathered at Stanford University to discuss the major needs for successful crop adaptation to climate change. To focus discussion over the three day period, the meeting centered on just three major crops – rice, wheat, and maize – given that these provide the bulk of calories to most populations. The meeting also focused on two aspects of climate– extreme high temperatures and extreme low moisture conditions (i.e. drought) – that present substantial challenges to crops in current climate and are likely to become more prevalent through time. Other aspects of climate change such as more frequent flooding or saltwater intrusion associated with rising sea levels were not addressed, although they may also be important.

The current document is split into two sections:

• a brief summary of material presented at the meeting on the current state of climate projections, crop modeling, crop genetic resources and breeding; and
  
• the collective views of participants on major needs for future research and investment, which emerged from discussions over the three day meeting.
  

The main target audiences for the document are donor institutions seeking to invest in climate adaptation, and climate and crop scientists seeking to set research agendas. We intend the term donor institutions to include private foundations, governments, and inter‐governmental organizations such as the World Bank and United Nations. An underlying assumption of the Stanford meeting was that there is a real and growing need to identify specific investment opportunities that will improve food security in the face of climate change. This is reflected, for instance, by the recent G8 announcement of a $20B investment in food security, the expectation of additional resources for adaptation from the Copenhagen Conference in 2009, and the emphasis of the Obama administration on food and climate issues.

This paper is part 2 of a two-part study evaluating the climatic effect of one of the nation's most rapidly expanding metropolitan complexes, the Greater Phoenix, Arizona, region.

Part 1, using a set of sensitivity experiments, estimated the potential impact of observed landscape evolution, since the dawn of the Landsat satellite era on the near surface climate, with a primary focus on the alteration of the surface radiation and energy budgets and through use of high-resolution, 2km grid spacing, Regional Atmospheric Modeling System (RAMS) simulations with circa 1973, circa 1992, and circa 2001 landscape data sets.

In this paper, part 2, we address the role of the previously discussed surface budget changes and subsequent repartitioning of energy on the mesoscale dynamics and thermodynamics of the region, the effect on convective rainfall, and their association with the large-scale North American Monsoon System (NAMS). Our results show that contrasts in surface heating resulting from landscape change are responsible for the development of preferentially located mesoscale circulations, on most days, which were stronger for the 2001 compared to the 1973 landscape, due to increased planetary boundary layer (PBL) heating via enhanced turbulent heat flux.

The effect of these stronger circulations was to warm and dry the lower part of the PBL and moisten the upper part of the PBL for the 2001 relative to the 1973 landscape. The precise physical pathway(s) whereby precipitation enhancement is initiated with evolving landscape, since the early 1970s, reveals a complicated interplay among scales (from the turbulent to the synoptic scale) that warrants future research. Precipitation recycling, however, was found to be an important driver in the overall sustenance of rainfall enhancement.

Although this study was not designed to investigate other radiative forcing factors such as greenhouse gas emissions and aerosols, the results of our sensitivity experiments do suggest that regional land use change is an important element of climate change in semiarid environments characterized by large urban areas with scarce water resources.

This chapter looks more qualitatively at Asia as a single region and focuses more speculatively on the longer run. It is also more personal, since views of the 2025, like beauty, depend inevitably on the eyes of the beholder. The chapter begins by running history backwards via a thought experiment. Suppose in 1975 that one had been predicting the contours of the Asian maize economy in 2000. What variables and parameters would forecasters have missed in significant ways? This procedure turns out to be a very sobering exercise for anyone now predicting events in 2025.

Biofuels from land-rich tropical countries may help displace foreign petroleum imports for many industrialized nations, providing a possible solution to the twin challenges of energy security and climate change. But concern is mounting that crop-based biofuels will increase net greenhouse gas emissions if feedstocks are produced by expanding agricultural lands. Here we quantify the 'carbon payback time' for a range of biofuel crop expansion pathways in the tropics. We use a new, geographically detailed database of crop locations and yields, along with updated vegetation and soil biomass estimates, to provide carbon payback estimates that are more regionally specific than those in previous studies. Using this cropland database, we also estimate carbon payback times under different scenarios of future crop yields, biofuel technologies, and petroleum sources. Under current conditions, the expansion of biofuels into productive tropical ecosystems will always lead to net carbon emissions for decades to centuries, while expanding into degraded or already cultivated land will provide almost immediate carbon savings. Future crop yield improvements and technology advances, coupled with unconventional petroleum supplies, will increase biofuel carbon offsets, but clearing carbon-rich land still requires several decades or more for carbon payback. No foreseeable changes in agricultural or energy technology will be able to achieve meaningful carbon benefits if crop-based biofuels are produced at the expense of tropical forests.

Rising populations and incomes throughout the world have boosted meat demand by over 75% in the last 20 years, intensifying pressures on production systems and the natural resources to which they are linked. As a growing proportion of global meat production is traded, the environmental impacts of production become increasingly separated from where the meat is consumed. In this paper, we quantify the use of three important resources associated with industrial livestock production and trade - water, land, and nitrogen - using a country-specific model that combines trade, agronomic, biogeochemical, and hydrological data. Our model focuses on pigs and chickens, as these animals are raised predominantly in intensive systems using concentrated, compound feeds. The results describe the geographical patterns of environmental resource use due to meat production, trade, and consumption. We show that US feed, animal, and meat destined for export require almost as much nitrogen and land, and 20% more water, than products destined for domestic consumption. Model results also demonstrate that among various production factors, improvements in crop yields and animal feed conversion efficiencies result in the most significant reductions in environmental harm. By explicitly tracking the externalities of meat production, we hope to bolster suppliers' accountability and provide better information to meat consumers.

Converting forest lands into bioenergy agriculture could accelerate climate change by emitting carbon stored in forests, while converting food agriculture lands into bioenergy agriculture could threaten food security. Both problems are potentially avoided by using abandoned agriculture lands for bioenergy agriculture. Here we show the global potential for bioenergy on abandoned agriculture lands to be less than 8% of current primary energy demand, based on historical land use data, satellite-derived land cover data, and global ecosystem modeling. The estimated global area of abandoned agriculture is 385-472 million hectares, or 66-110% of the areas reported in previous preliminary assessments. The area-weighted mean production of above-ground biomass is 4.3 tons/ha^-1 /y^-1, in contrast to estimates of up to 10 tons/ha/yr in previous assessments. The energy content of potential biomass grown on 100% of abandoned agriculture lands is less than 10% of primary energy demand for most nations in North America, Europe, and Asia, but it represents many times the energy demand in some African nations where grasslands are relatively productive and current energy demand is low.

» Article in the Stanford Report on Campbell et al. 
» Video by the Stanford News Service.