H11I-01 INVITED
Flood Flows and Climate Variability and Change in the US, an Exploration of the Literature, Theory, and Long-term Flood Records
Much of the recent literature about the relationship of greenhouse-gas induced warming and extreme precipitation has indicated that extreme events, in many parts of the U.S. and globally, have become more common over time. Further, global climate models predict that extreme precipitation events will become more common as greenhouse gas concentrations rise. Logic would suggest that flood flows would have increased in recent decades in response to increases in the frequency of extreme precipitation. Research to date, however, is highly equivocal on the question of increasing flood magnitudes (as distinct from economic flood damages which have shown a strong increase). Emergency preparedness and mitigation of flood damages in the future should be based on a strong scientific foundation of understanding of the reasons for the changes to date (if any) and the likely drivers of change in the future. Some of the drivers of change are undoubtedly related to land use and water management strategies. The question of a climate driver also needs exploration: including analysis of historical relationships between flood flows and greenhouse gas concentrations, as well as relationships to various indices of ocean conditions that are known to be major drivers of the hydro-climatic system at multi- year to multi-decadal time scales. Additionally, climate models need to be critically evaluated in terms of their current usefulness for estimating flood behavior at various basin scales and in various climatic zones. This study will review the literature and theory behind this issue and explore a small number of the very longest U.S. flood records to search for trends and possible relationships between flood magnitude and climate forcing.
H11I-02
Changes in Monthly Streamflow Conditions in the Missouri River Basin from 1957 to 2007
Observations of 50 years of continuous record at about 200 U.S. Geological Survey gages indicate that streamflow conditions are changing in the Missouri River Basin (MRB). Trends are evident in the annual streamflow records at 81 stations using the non-parametric Kendall Tau test. Where trends are significant, they generally are upward in the eastern portions of the basin and downward in the western portions of the basin. The reduced runoff in the western basin has resulted in main-stem reservoirs on the Missouri River that have operated at much less than full capacity for most of the last decade. Lower reservoir capacities result in significant economic effects such as reduced hydropower revenues, reduced recreation opportunities, and lower basin barge traffic. A change in the timing of streamflow or seasonality merits careful examination due to the implications for reservoir management, water supply strategies, and ecological ramifications. In this study, we examine streamflow trends by month within the MRB for the period 1957 to 2007. Two data sets were examined—202 stations with 51 years of record and a subset of 81 stations that presented a significant trend in annual flow. For both data sets, the month of April has the most stations (66) with upward trends, followed by March (59), then February (44) and May (44). The month of June has the fewest stations (18) with downward trends. Similarly, for both data sets, the months of September (57) and December (56) have the most stations with downward trends. The month of August has the fewest stations with downward trends (21). There is a strong geographic clustering of stations with trend directions. For example, most stations with trends in Colorado, North Dakota, South Dakota, Iowa, and Missouri are upward, whereas most stations with trends in Montana and Wyoming are downward. In 81 percent of the cases, more than one month exhibited a trend for a given station. The amount of water that these upward or downward trends represent will be presented by month. A statistically significant trend for a given month may, in some cases, represent a relatively small amount of runoff.
H11I-03 INVITED
Global and Regional Variations and Trends in Precipitation and Relations to Surface Temperature Changes
Associations between rainfall and surface temperature anomalies on the longer-than-seasonal time scales are explored for the period of 1979-2007, using the GPCP precipitation product and the GISS surface temperature data set. Positive (negative) correlations are generally observed between these two variables over tropical oceans (lands) on an inter-annual basis. ENSO is the dominant factor in these inter-annual, tropical relations. In the northern hemisphere mid-to-high latitudes, the correlation relationships between rainfall and temperature anomalies are more complicated with positive and negative values of correlation tending to appear over both oceans and land. Furthermore, there seems a strong seasonal variation in correlation in the north hemisphere middle and high latitudes. These relations are due to a complex combination of ocean and land local effects, seasonal variations and dynamics. The main long-term, linear change during the period is in the deep tropics with decreases in the sub-tropics. For surface temperature, however, both the strongest linear changes and largest variances are observed in the northern hemisphere mid-high latitudes, with much weaker temperature changes in the tropical region and southern hemisphere. The long-term, zonally-averaged precipitation change to temperature change resembles the inter-annual version of the same ratio, except in the tropics over land, indicating some overlap in processes affecting the relations on the two time scales.
H11I-04
ENSO's Global Influence on the Seasonal Distribution of Daily Rainfall
Hazard, flood and landslide applications, especially in developing countries, have prompted the recent examination of changes to extreme rainfall using global satellite information. However, few studies have investigated changes to the entire distribution of rainfall, which is important for many hydrological applications. Interannual variability, in particular, can be determined with a two-sample non-parametric Kolmogorov-Smirnov (KS) test. The KS-test is computed between pairs of years on all available daily 0.25 x 0.25 degree Tropical Rainfall Measuring Mission Multisatellite Precipitation Analysis (TMPA) grid boxes seasonally. Then the fractional coverage of significant differences is related to the absolute difference of Nino 3.4 between the two years for the concurrent season and previous seasons. It was found that ENSO explains a large portion of the global variance, with the recent El Nino and La Nina events demonstrating atypical distributions of daily rainfall. Finally, this paper begins to point to regions where ENSO's influence on seasonal rainfall may be different than its impact on weather statistics within a season, helping to direct data gathering efforts for comprehensive regional studies.
H11I-05
Emerging Trends in Terrestrial and Global Water Storage from GRACE
The GRACE time-variable gravity mission has provided a new (since 2002) and unique opportunity to observe changes in terrestrial and global water storage from space. As the mission approaches seven years in orbit, we can now begin to characterize interannual variations in total water storage on land, water mass variations in Earth's land, ocean, atmosphere and ice reservoirs, and emerging trends from river basin to continental and global scales. In this presentation we discuss the latest trends on land and in the global water cycle, with implications for water cycle acceleration and for fresh water availability on land.
H11I-06
Evidence for Changing Flood Risk in New England Since the Late 20th Century
Long-term flow records for watersheds with minimal human influence have shown trends in recent decades toward increasing streamflow at regional and national scales, especially for low flow quantiles like the annual minimum and annual median flows. Trends for high flow quantiles are less clear, despite recent research showing increased precipitation in the conterminous United States over the last century that has been brought about primarily by an increased frequency and intensity of events in the upper 10th percentile of the daily precipitation distribution, particularly in the Northeast. This study investigates trends in 28 long-term annual flood series for New England watersheds with dominantly natural streamflow. The flood series are an average of 75 years in length and are continuous through 2006. Twenty-five series show upward trends via the non-parametric Mann-Kendall test, forty percent (10) of which are statistically significant (p<0.1). Moreover, an average standardized departures series for 23 of the study gages indicates that increasing flood magnitudes in New England occurred as a step change around 1970. The timing of this is broadly synchronous with a phase change in the low frequency variability of the North Atlantic Oscillation (NAO), a prominent upper atmospheric circulation pattern that is known to effect climate variability along the United States east coast. Identifiable hydroclimatic shifts should be considered when the affected flow records are used for flood frequency analyses. Special treatment of the flood series can improve the analyses and provide better estimates of flood magnitudes and frequencies under the prevailing hydroclimatic condition.
H11I-07
Long-Term Hydrological Changes (1946-2006) of the Seine River Discharge, France
The variability of Seine river discharge (France) was investigated using daily time series over the 1946-2006 period. The study focuses on the determination of the dominant modes which control the variability of discharge and precipitations in the Seine river watershed and of their possible link with the North Atlantic Oscillation. Overall, the Seine river discharge is affected by a statistically significant increasing trend across the period. The hydrologic regime of the Seine river was found to be highly variable, especially during high-water periods (fall and winter), which seem to occur later in the year for approximately 15 years. The maximum annual discharge time series exhibited an obvious change around 1970, while the minimum annual discharge is affected by a change rather around 1990. Continuous wavelet analysis of precipitations and discharge revealed similar spectral patterns in the form of energetic peaks highly localized in time, defining three time periods: before 1970, between 1970 and 1990, and after 1990. The discontinuities found in the extreme discharge values were then recovered. Overall, a 17-yr mode occurs around 1970 and a 6.4-yr shows up in the late 1980. The annual oscillation is always strongly represented in discharge, although it is affected by an increase in power throughout the period of study, much more powerful from 1990 until the end of the series. The daily NAO wavelet transform is characterized by 1-yr and 2.8-3.7-yr fluctuations which loses power across the period of study and by a lower 8.3-yr interannual occuring around 1970. In the late 1980, the NAO 2.8-3.7-yr interannual shifts towards slightly highest frequencies. According to their wavelet spectra, hydrometeorological data and NAO did not seem to oscillate over the same modes of variability, although similar temporal discontinuities were revealed. However, the calculation of the NAO/Seine discharge wavelet coherence emphasized a strong degree of linear correlation (coherence close to 1) between the two signals at annual time scales starting from the late 1980's. This result rises up the question of an intensification of the hydrologic cycle in the Seine river watershed in relation with the strongly positive phases of NAO from the very end of the last century up to now.
H11I-08
Estimation of permafrost thawing rates in a sub-arctic catchment using recession flow analysis
Permafrost may be a good indicator of climate change. Permafrost thawing is likely to change the flow pathways taken by water as it moves through arctic and sub-arctic landscapes. The location and distribution of these pathways directly influence the biogeochemistry of terrestrial water and changes in the major pathways affect carbon and other biogeochemical cycling in northern latitude catchments. Direct observations of depth to permafrost are difficult to perform at scales larger than a local scale. Using recession flow analysis, it should be possible to estimate the rate of permafrost thawing based on a long-term streamflow record. The strength of such an approach is that it uses hydraulic theory to infer changes in depth to permafrost at the support scale of the catchment. We demonstrate this approach for the sub-arctic Abiskojokken catchment in northern Sweden. Based on recession flow analysis, we estimate the permafrost in this catchment to be thawing at a rate of about 0.9 cm per year during the past 90 years. This estimate is supported by 30 years of direct observation in the region.