Taking a Systemic Look at Char...

Taking a Systemic Look at Characteristics of the Global Hydrologic Cycle

Changing climate will directly affect the global hydrologic cycle. Many of these effects will be felt regionally, with, for example, potential for flooding or drought increasing. In addition, changes to water quality, quantity, and supply reliability may have effects on human health, aquatic ecosystems, and agricultural and energy production, among other ecosystems and economic sectors. Recognizing a growing need to better understand climate change effects on regional water cycles around the world, the National Center for Atmospheric Research’s (NCAR) Water System Program strives to advance hydrologic-cycle research and modeling. Today’s climate models agree fairly well on temperature projections, however future projections of precipitation – along with other components of the water cycle, such as evapotranspiration and runoff– are notoriously inconsistent between models and especially so at regional scales. 

A cross-NCAR and cross-university effort, Water System Program leaders have chosen a few key areas in which to focus research efforts. Among these:

  • Diagnosing the global water cycle behavior in models using observations, data sets, and model output.

  • Study key processes controlling regional and global precipitation and use this knowledge to improve convective and orographic precipitation parameterizations.

  • Investigate possible futures of the water cycle in the U.S. West under changing climate conditions

  • Improve simulation of snowpack conditions in climate models by comparing SNOTEL observations to high-resolution model simulations, and by conducting a snow physics inter-comparison project.

  • Examine the impact of land-surface disturbances on the water cycle.

  • Analyze the effects of water-resource management (e.g., dams, irrigation) on the Earth’s systems.

Predicting Drought

Among the efforts to improve drought prediction is a study by the Water System Program’s Aiguo Dai on understanding climate models’ capabilities for accurate prediction of future drought patterns in the United States. When comparing real-world observations of historical precipitation, streamflow, and drought patterns to modeled predictions of drying trends generated by climate models, modeled and observed patterns often do not match. One reason for this is that climate models typically have difficulty with accurately replicating air-ocean interactions, which in turn affects projections of continental precipitation patterns. Climate scientists often inferred that this mismatch meant that climate models were over-predicting the probability of future drought-event occurrences and severity.

To address climate models’ difficulties in accurately replicating sea surface temperatures, Dai developed a statistical method to account for the natural variation in sea surface temperature. By using data that took this variation into consideration, Dai found that climate models in fact were reproducing past climate conditions better than realized, meaning that climate models’ projections about the probability of future drought are likely accurate. One worrying upshot of this study is that in future projections, model output indicates that the world will likely see more severe and widespread droughts in the next 30-90 years in regions of North and South America, Africa, and southern Europe under changing climate conditions. 

Understanding Effects of Climate Change on Snowmelt-Dependent Water Basins

A long-term focus of members of the Water Systems Program has been modeling the Colorado River Basin’s flow under changing climate conditions. Like many other river basins in the world, the Colorado River’s hydrologic system gets much of its moisture in the form of seasonal snow melt. To understand the effects of future climate on the Basin, the team worked to refine their understanding of the river’s water-cycle dynamics, while at the same time honing the modeling tools to provide finer scale, more precise information. In doing so, the updated models show that future changes in winter precipitation, combined with increasing rates of evapotranspiration (the amount of water that evaporates or is used by plants as they grow) may, in future, reduce the amount runoff within the Colorado Basin even under climate conditions that increase the snowfall in the region. With both farmers and urban regions in the U.S. West reliant upon these resources, this is critical information for regional water-resource users and planners. In addition to the regional benefits, the modeling lessons learned based on the Colorado River Basin will be transferred to study and develop an improved understanding of other hydrologic systems around the world (e.g., the Himalayan hydrologic system).

Pine Beetles, Climate Change, and the Hydrologic Cycle

Bark beetle infestation across western North American forests has caused the death of millions of acres of pine trees over the past decade. The massive die-off has left scientists wondering about the possible effects on the hydrologic cycle. A three-year study, which focused on this question and included scientists from the Water System Program and NCAR’s university community, ended in 2012. Early research results are coming in that shed insights on changes in forest ecosystems. Most researchers have assumed that in areas in which large swaths of trees are dead or dying due to beetle infestation, the hydrologic cycle would experience increased runoff, higher soil-moisture content, and, in winter, increased peak snowpack conditions. Instead, the scientists are finding that things may not be so straightforward and that while the affected pine trees may not be using as much water, all the surplus water may not be finding its way into streams and rivers. Other plants underneath the dead and dying trees or in adjacent communities seem to be taking up the excess available water coming into the system.  Also, more water seems to be evaporated from the forest soils and sublimated (when water goes from ice to vapor phases) from the snowpack on the ground compared to when the forest canopy is full and shading and sheltering the ground from the sun and wind. Also, year-to-year climate variability – a drought year, followed by a year of heavy snow, followed by low snowpack and cool summer temperatures, for example – seems to be having an overall greater impact on these pine-forest ecosystems than are the effects of beetle kill, making it difficult to quantify the exact impacts of the dying stands of trees.

“An emerging idea from this research is how poorly we understand an ecosystem’s compensating effects on the water cycle in regions of beetle kill,” says David Gochis, a researcher in NCAR’s Research Applications Laboratory and member of the Water Systems Program. “Our ability to see these effects may be a function of scale. At stand scales (several tens to hundreds of meters or less), the compensating effects are more difficult to see than is the case at a larger scale when growth of new vegetation or of non-impacted vegetation uses the available runoff that the pine trees would have used previously.”

Intercomparison of Snowpack Models Covering the Complex Terrain of the Forested Central Rocky Mountains

Spring snowmelt runoff timing and amount in mountainous regions are critical pieces of information for water resources management. An effort is underway within the university community to utilize data from 112 SNOTEL sites in the Colorado Headwaters region and two Ameriflux sites to evaluate the ability of six widely-used land-surface/snow models (Noah, Noah-MP, VIC, CLM, SAST, and LEAF-2) in simulating the seasonal evolution of snowpack in the central Rockies.

All models captured seasonal evolution of snow water equivalent (SWE; the amount of water equivalent in a snow pack) fairly well (Figure 3). However, they underestimated both early-spring (March-April) snow accumulation and late spring ablation. Underestimating snowmelt from mid-May to mid-June allowed models to compensate for lower SWE estimates in spring, and consequently resulted in each showing a prolonged snow season. No single model excelled at (or fell overly short of) reproducing the three important features of snow evolution: maximum SWE depth, the timing of maximum SWE, and the timing of spring snow disappearance.

However, models exhibited large disparities in simulating the surface energy partitioning, which is equally important for correctly representing snow-atmospheric interactions in weather and climate models. Some models underestimated the solar energy absorbed at the forest-soil-snow interface from December to March. That resulted in the models showing too little outgoing long-wave radiation and sensible heating being returned to the atmosphere, which could be a crucial deficiency for coupled weather and climate models. Those model disparities and deficiencies can be further traced down by examining the treatment (or lack of treatment) of turbulence and radiation processes within and under the vegetation canopy. Excessive shortwave and long-wave radiation transfer from canopy to the ground/snow often led to larger than actual sensible heating (i.e., the energy required to change the temperature of a substance) from canopy to ground/snow surface, and resulted in larger, undesired weekly SWE change in both accumulation and ablation phases. Accurate radiation transfer between canopy and ground/snow is essential for capturing both snowpack evolution and snow-vegetation-atmosphere interactions.