Terrestrial Hydrology Science: Storage Change and Discharge. D. Alsdorf, P. Bates, A. Boone, F. Hossain, T. Pavelsky, and Y. Sheng

Transcription

1 SWOT Terrestrial Hydrology Science: Storage Change and Discharge D. Alsdorf, P. Bates, A. Boone, F. Hossain, T. Pavelsky, and Y. Sheng funded by: CNES, JPL, NASA, and OSU s Climate, Water, & Carbon Program D. Alsdorf March, 2010

2 Amazon Water Balance S = zero = P E Qs m/yr mm/day m 3 /s m 3 /s E from Costa, M.H., and J.A. Foley, JGR, 1999; P from Costa, M.H., and J.A. Foley, GRL, 1998.

3 Global Water Cycle & Climate Modeling Runoff (mm/day) Observed Models How does the lack of measurements limit our ability to predict the land surface branch of the global hydrologic cycle? In locations where gauge data is available, GCM precipitation and subsequent runoff miss streamflow by 100% The question is unanswered for ungauged wetlands, lakes, and reservoirs throughout the world. Roads et al., GCIP Water and Energy Budget Synthesis (WEBS), J. Geophysical Research, 2003

4 Photo: Larry Smith

5 1. 2. Overview: International River Basins (IRB) About large artificial reservoirs in the US (~approx of 1 year of annual Tennessee runoff impounded). In the Technological world every other river is impounded (World Commission on Dams). Large reservoirs are intimately impacted by humans, national interests and transboundary issues. University % Area of a Country inside an IRB Number of Countries 91-99% % % % % % % % % % 11 Tennessee Tech UNIVERSITY Map showing the density of large dams in international river basins in Asia. Courtesy: Dr. Aaron Wolf, Oregon State

6 6 Carlisle, UK 10m model vs. ground survey RMSE on water depth = 0.32 m

7 Long Version

8 SWOT Terrestrial Hydrology Science: Storage Change and Discharge D. Alsdorf, P. Bates, A. Boone, F. Hossain, T. Pavelsky, and Y. Sheng funded by: CNES, JPL, NASA, and OSU s Climate, Water, & Carbon Program D. Alsdorf March, 2010

9 2.2.0 Introduction Credit: CNES Global Water Cycle Aaron Boone Arctic Hydrology Tamlin Pavelsky Lakes Yongwei Sheng Reservoirs and Human Impacts Faisal Hossain Floodplain Processes Paul Bates Flooding Paul Bates

10 Terrestrial Water Balance S = P E Qs Qg Applied to an entire basin: Key = time S: Storage change summed from soil moisture, snow water content, surface water storage, vegetation water content, ground water, and glaciers P, E: Precipitation and Evaporation Qs: River discharge Qg: Groundwater flux across basin boundary

11 Amazon Water Balance S = zero = P E Qs m/yr mm/day m 3 /s m 3 /s E from Costa, M.H., and J.A. Foley, JGR, 1999; P from Costa, M.H., and J.A. Foley, GRL, 1998.

12 Amazon Downing et al., 2006 Siberia Ohio Photos: K. Frey, B. Kiel, L. Mertes

13 Photo: Larry Smith

14 Lakes in Peace-Athabasca, Alaska, and Siberia. Approach uses altimetry, remote sensing, in-situ, and statistics to create truth. SWOT orbits and height errors to create storage change estimates.

15

16 SWOT capability to estimate discharge Ohio River accuracy estimate based on: 1) Expected SWOT height observation accuracy 2) Temporal sampling errors 3) River width Courtesy: Elizabeth Clark, UW

17 SWOT Simulated Discharge Along the Kanawha River Durand et al., 2009

18 2.2.0 Introduction Credit: CNES Global Water Cycle Aaron Boone Arctic Hydrology Tamlin Pavelsky Lakes Yongwei Sheng Reservoirs and Human Impacts Faisal Hossain Floodplain Processes Paul Bates Flooding Paul Bates

19 Global Water Cycle & Climate Modeling Runoff (mm/day) Observed Models How does the lack of measurements limit our ability to predict the land surface branch of the global hydrologic cycle? In locations where gauge data is available, GCM precipitation and subsequent runoff miss streamflow by 100% The question is unanswered for ungauged wetlands, lakes, and reservoirs throughout the world. Roads et al., GCIP Water and Energy Budget Synthesis (WEBS), J. Geophysical Research, 2003

20 Model Improvements Despite recent improvements, the aforementioned river routing parameterizations use very simple assumptions, largely owing to parameter uncertainty and a lack of high quality spatially distributed evaluation data, such as discharge, over large parts of the globe. Moreover, it has been shown by inter-comparing a large ensemble of independent LSMs driven by realistic atmospheric forcing that on a global scale, the land surface flux with the highest uncertainty is the runoff (i.e., least agreement defined as the largest inter-model variabilitydirmeyer et al., 2006). SWOT would provide the first possibility to improve the river routing, the representation of lakes and hydrological model parameterizations within LSMs at a relatively high spatial resolution over the entire globe using a data assimilation strategy. Tendencies in water height change and spatial coverage could be directly assimilated into river routing and floodplain parameterizations, which would then result in improved estimates of discharge on a global scale. The more realistic estimates of river routing could then be used to improve LSM runoff parameterizations and, by extension, estimates of evaporation.

21 21 SWOT & GCM-RCM Continental Hydrology ISBA-TRIP coupling system Land surface model (ISBA) model to generate runoff River routing using TRIP (1 or 0.5 degrees) SWOT, OSU, Sept

22 22 SWOT & GCM-RCM Continental Hydrology The ISBA-TRIP coupling system Key scientific question for SWOT: better quantify the echange between rivers and floodplains for improved prediction TRIP Key variables from SWOT ISBA Development Methodology 1) Improve in «offline» mode validation with obs and/ or assimilation 2) Use in (fully-coupled) projection extrapolate in n,t! * NEED Global scale data! SWOT, OSU, Sept

23 Key Point for Global Water Cycle Science: Global water cycle modeling fails to adequately define surface water storages and fluxes, yet the these are critical to improving model performance and hence our ability to predict future water resources under conditions of changing climate and anthropogenic forcings. Measurements: The main need for SWOT products in terms of GCM applications is to provide multi-year data at the global scale of discharge, water height and slope within the context of improving the hydrological parameterizations: notably in terms of exchanges between rivers and floodplains, changes in lakes/ wetlands, and freshwater discharge into the oceans.

24 2.2.0 Introduction Credit: CNES Global Water Cycle Aaron Boone Arctic Hydrology Tamlin Pavelsky Lakes Yongwei Sheng Reservoirs and Human Impacts Faisal Hossain Floodplain Processes Paul Bates Flooding Paul Bates

25 Where are the World s Lakes? Published databases suggest that rivers and lakes north of 55 N represent: >30% of global open water areas >50% of all lakes larger than 0.1 km 2 Gauging even 1% of these lakes in situ is unfeasible but SWOT can track all of them. Lehner and Doll, 2004 (Global Lakes and Wetlands Database)

26 The current lake & wetland extent is poorly known, let alone storage. West Siberia Frey and Smith, GBC, 2008

27 Smith et al., Disappearing Arctic Lakes, Science, 2005 Arctic lakes are losing storage, despite a slight increase in precipitation. The spatial pattern of lake loss strongly suggests that the melting of permafrost is driving the process (rather than evaporation). At first, permafrost melting increases lake storage, but continued melting breaches the underlying frozen ground allowing the lake to drain into the subsurface.

28 Braided Rivers Permafrost Courtesy International Permafrost Assn. Arctic Rivers Among the most intractable problems in Arctic hydrology over the last decade is determining the source of recently observed increases in river discharge into the Arctic Ocean. SWOT will: Track discharge in braided rivers that cannot be easily gauged in situ. Allow direct assessment of the effect of permafrost extent on downstream discharge variation. Monitor flow out of glacial rivers, tracking seasonal melt patterns and observing large glacial outburst events. Observe river-floodplain interactions during the critical spring ice-breakup season, when most water and nutrient exchange occurs. Courtesy Stephen Codrington Glacial Rivers River Ice Breakup

29 Arctic Hydrology The ecological function of floodplains and deltas, among the most biologically productive environments in the Arctic, is strongly dependent on recharge of water and sediment from rivers. There is substantial concern regarding drying of Arctic floodplain wetlands due to decreases in spring flooding associated with dam construction and climate change. Arctic lakes represent a substantial source of atmospheric methane (CH 4 ) that may be sensitive to lake extent variations associated with changing permafrost conditions. However, estimates of total contemporary methane flux from Arctic lakes remain uncertain due to poorly characterized estimates of inundated area. Moreover, changes in surface water hydrology may also result in increased emissions of CO 2 to the atmosphere and dissolved organic carbon into the Arctic Ocean from Arctic peatlands. Improved knowledge of mean inundation extent and seasonal dynamics in water level from SWOT will decrease uncertainty in current fluxes of CH 4 and CO 2.

30 Prairie Pothole Region Summation of small lakes S is an indicator of P, E, and groundwater, i.e., drought and deluge.

31 Key Point for Arctic Hydrology and Global Lakes Science: Arctic surface water hydrology is linked to cryospheric processes associated with ice sheets, glaciers, river and lake ice, and permafrost. Climate change and water cycle acceleration expressed in these linkages as changes in Q and S. Impacts on carbon balance from these changes is not known, e.g., inundation of floodplains and the related exchange of carbon and nutrients is not well known. Measurement: The abundance of Arctic lakes suggests that S is a key hydrologic driver, yet is unknown. By allowing substitution of contemporary, permafrost-driven spatial variations in hydrologic regime (i.e., S and Q) for future temporal changes in permafrost extent, SWOT observations will improve projections of climate change impacts on Arctic lake hydrology.

32 2.2.0 Introduction Credit: CNES Global Water Cycle Aaron Boone Arctic Hydrology Tamlin Pavelsky Lakes Yongwei Sheng Reservoirs and Human Impacts Faisal Hossain Floodplain Processes Paul Bates Flooding Paul Bates

33 1. 2. Overview: International River Basins (IRB) About large artificial reservoirs in the US (~approx of 1 year of annual Tennessee runoff impounded). In the Technological world every other river is impounded (World Commission on Dams). Large reservoirs are intimately impacted by humans, national interests and transboundary issues. University % Area of a Country inside an IRB Number of Countries 91-99% % % % % % % % % % 11 Tennessee Tech UNIVERSITY Map showing the density of large dams in international river basins in Asia. Courtesy: Dr. Aaron Wolf, Oregon State

34 Key Point for Reservoirs and Human Impacts Science: The knowledge of storage and spillway discharge is controlled by nations in which the reservoir is located. No adequate treaties exist for trans-boundary cooperation at operational time scales. Controlling nation has no legal obligation to share reservoir information with downstream nations in timely manner. Thus: 1. Will SWOT data availability solve the fundamentally intractable problems of forecasting of trans-boundary water flow? 2. Will nations in IRBs become more and more independent and sovereign in forecasting and managing (impounded) water resources flowing from and to other nations? 3. Will the increased transparency of information increase trust among nations for greater cooperation on transboundary water issues? Measurement: Storage changes in reservoirs and the linkages of these to downstream flows, preferably on operational timescales.

35 2.2.0 Introduction Credit: CNES Global Water Cycle Aaron Boone Arctic Hydrology Tamlin Pavelsky Lakes Yongwei Sheng Reservoirs and Human Impacts Faisal Hossain Floodplain Processes Paul Bates Flooding Paul Bates

36 CO 2 Evasion from Water to Atmosphere What is the role of wetland, lake, and river water storage as a regulator of biogeochemical cycles, such as carbon and nutrients? e.g., Rivers outgas as well as transport Carbon. Ignoring water borne C fluxes, favoring landatmosphere only, yields overestimates of terrestrial C accumulation Amazon Floodplain Results: 470 Tg C/yr all Basin; 13 x more C by outgassing than by discharge. But what are seasonal and global variations? If extrapolate Amazon case to global wetlands, = 0.9 Gt C/yr, 3x larger than previous global estimates; Tropics are in balance, not a C Sink? Richey, J.E., J.M. Melack, A.K. Aufdenkampe, V.M. Ballester, and L.L. Hess, Outgassing from Amazonian rivers and wetlands as a large tropical source of atmospheric CO2, Nature, 416, , 2002.

37 That was a very coarse resolution, what about high resolution? How does water flow through this floodplain? Which channels convey the most water? Where does water reside the longest?

38 Conventional Idea of Floodplain Inundation

39 Measurements of Floodplain Inundation Localized, complex patterns of dh/dt Sharp dh/dt aligned with many channels Purus flood wave is apparent

40 40 Carlisle, UK 10m model vs. ground survey RMSE on water depth = 0.32 m

41 41 Upton on Severn, UK 18m model vs Airborne SAR Model fit = 89% Hydraulic models for flood prediction need extent images and water elevations to calibrate friction parameters and provide independent validation = correct = over-prediction = under-prediction = predicted as flooded, no ASAR coverage

42 Key Point for Floodplain Processes and Flooding Science: 1. How does inundated area vary annually across the globe and how do these spatial and temporal variations impact processes such as methane emission? 2. What are the residence times and flow paths of floodplain flow and what are the implications for spatial patterns of biogeochemical cycling, ecology, and sediment transport? 3. Where in the floodplain does terre-firme runoff mix with river water and how does this mixing of waters with different chemical and ecological signatures change dynamically? 4. What volume of water is exchanged between rivers and their floodplains, and how does this storage and release of water affect the downstream propagation of the flood pulse? Does this floodplain storage (which is entirely ungauged) lead to an underestimation of global terrestrial runoff? 5. How do floodplain and wetland flows interact with complex topography, vegetation and buildings, and how do the storage effects and energy losses so generated control the development of hazardous flooding? Measurement: SWOT will address the above questions either directly, through measurements of inundation extent, water surface elevation, h, its temporal and spatial derivatives dh/dt and dh/dx, S, and Q, or indirectly, by better constraining models of the above processes. Furthermore, a SWOT mission byproduct will also be the first global DEM of floodplain and wetland area with decimetric accuracy that can at last be used to parameterize hydraulic and hydrologic models of these systems adequately.

43 SWOT There are a host of scientific and societal questions that will be directly addressed by SWOT s hydrodynamic measurements and Q and S products. Questions range from impacts on the water cycle from climate change to flooding hazards in developed floodplains. Surface water flow is 2-dimensional, spatially. SWOT will match this with 2D measurements. This is a revolutionary change from our current 1D streamgauging approach. This mission is for everybody, please join us via: swot.jpl.nasa.gov/ D. Alsdorf March, 2010

44 Extra Slides

45

46