1 Modeling the synergistic impacts of atmospheric and land-based influences on water quality and quantity in linked upland and estuarine ecosystems Zong-Liang Yang (PI), David Maidment (co-PI), Paul Montagna (co-PI), James McClelland (co-PI), Hongjie Xie (co-PI) Guo-Yue Niu, Seungbum Hong, Cédric David, Hae-Cheol Kim, Sandra Arismendez, Rae Mooney, Patty Garlough, Rachel Mills, Beibei Yu, Ling Lu, Almoutaz El Hassan May 2010

Outline Introduction Models and tools Climate change in Texas Noah land surface model with multi-physics options River network model (RAPID) Statistical modeling of nutrient outflow Estuary model Summary and accomplishments New IDS 2

Introduction: An incomplete water cycle Atmosphere  Land –To atmospheric scientists: evapotranspiration (ET), space/time distribution; runoff immediately lost to ocean, treated as a residual of P – ET –To hydrologists: runoff, space/time; river routing, ET is a residual (P – Runoff) Land  Oceans –To terrestrial hydrologists: figures mask oceans –To marine scientists: figures mask land 3

Enter the coastal environments! 4

Zong-Liang Yang (UT Geological Sciences) Climate modeling Land surface modeling Hongjie Xie (UTSA Geological Sciences) Remote sensing analysis LULC, NEXRAD David Maidment (UT Civil Engineering) Hydrology, river routing Land/ocean coupling, BGC cycles Jim McClelland (UT Marine Science) Paul Montagna (TAMU Corpus Christi) Estuary ecology Coastal BGC Modeling framework

6 Framework for calculation Atmospheric Model or Dataset Vector River Network - High-Performance Computing River Network Model Land Surface Model Modeling across spatial and temporal scales: Global  regional  watershed  coastal Current and future: interannual to hourly

Area of interest for integration with bays 7 San Antonio Bay Mission Bay, Copano Bay and Aransas Bay

Guiding questions What is the effect of global climate change on complex coastal environment processes? What are the quantities of nutrient and organic matter export from land to sea? How does the timing of export events influence ecosystem structure and productivity in coastal waters? How do changes on land influence watershed export and ecosystem properties in coastal waters? 8

Precipitation gradient across Texas Texas is home to four of 10 largest cities in the nation. Chemical plants and oil refineries are located along the coast near Houston. Precipitation increases from semi-arid west to humid east. Temperature decreases from south to north.

Precipitation comparison Texas double peaks of monthly precipitation: spring, fall

Models and tools 12

Weather models and datasets 13 WRF model Climate models NASA datasets (NLDAS, satellite datasets of land cover and change) Radar observations NEXRAD data

14 Land Surface Models for Land – atmosphere processes First version in 1999 Noah is fully coupled with WRF in North America Noah model

15 NHDPlus – River and Catchment Network for the Nation 3 million river reaches Integration of the National Hydrography Dataset, National Elevation Dataset and National Land Cover Dataset completed by EPA in ,175 river reaches 26,000 km 2 Guadalupe and San Antonio Basins, TX Entire dataset Connectivity information

River gages and dam operation data 16 Historical and real-time measurements Dam operation data

Parallel computing for river flow mapping 17 Desktop computer Parallel computer (UT’s Lonestar) Today’s computers are as powerful as supercomputers ten years ago Most computers come with multiple processors High performance parallel computers are becoming increasingly accessible

Climate change in Texas 18

Data WCRP CMIP3 dataset 16 global climate models Three emission scenarios (A1B, A2, B1) A1B39 simulations A237 simulations B136 simulations Precipitation and temperature (monthly) Statistical downscaling with bias-correction

Temperature projections

Projected probability distributions of surface temperature changes in the period of 2070–2099 relative to means by different climate models over Texas

Projected precipitation changes

Projected monthly precipitation anomalies over individual five regions in Texas

Summary Texas is getting warmer (2-5 º C by the end of this century); more warming in the north than in the south. Overall decreasing trend of precipitation. Decreasing precipitation in the winter (5-15%), and increasing precipitation in the summer (5%). We have downscaled precipitation (and other variables) at 3-hourly and fine-spatial scales for hydrological studies. Bias correction is applied to precipitation before it is used to drive hydrological models.

Noah land surface model with multi-physics options 25

Noah-MP with multi-physics options 1. Leaf area index (prescribed; predicted) 2. Turbulent transfer (Noah; NCAR LSM) 3. Soil moisture stress factor for transpiration (Noah; BATS; CLM) 4. Canopy stomatal resistance (Jarvis; Ball-Berry) 5. Snow surface albedo (BATS; CLASS) 6. Frozen soil permeability (Noah; Niu and Yang, 2006) 7. Supercooled liquid water (Noah; Niu and Yang, 2006) 8. Radiation transfer: Modified two-stream: Gap = F (3D structure; solar zenith angle;...) ≤ 1-GVF Two-stream applied to the entire grid cell: Gap = 0 Two-stream applied to fractional vegetated area: Gap = 1-GVF 9. Partitioning of precipitation to snowfall and rainfall (CLM; Noah) 10. Runoff and groundwater: TOPMODEL with groundwater TOPMODEL with an equilibrium water table (Chen and Kumar, 2001) Original Noah scheme BATS surface runoff and free drainage More to be added Niu et al. (2010a,b)

Maximum # of Combinations 1. Leaf area index (prescribed; predicted) 2 2. Turbulent transfer (Noah; NCAR LSM) 2 3. Soil moisture stress factor for transp. (Noah; BATS; CLM) 3 4. Canopy stomatal resistance (Jarvis; Ball-Berry) 2 5. Snow surface albedo (BATS; CLASS) 2 6. Frozen soil permeability (Noah; Niu and Yang, 2006) 2 7. Supercooled liquid water (Noah; Niu and Yang, 2006) 2 8. Radiation transfer: 3 Modified two-stream: Gap = F (3D structure; solar zenith angle;...) ≤ 1-GVF Two-stream applied to the entire grid cell: Gap = 0 Two-stream applied to fractional vegetated area: Gap = 1-GVF 9. Partitioning of precipitation to snow- and rainfall (CLM; Noah) Runoff and groundwater: 4 TOPMODEL with groundwater TOPMODEL with an equilibrium water table (Chen and Kumar, 2001) Original Noah scheme BATS surface runoff and free drainage Niu et al. (2010a,b) 2x2x3x2x2x2x2x3x2x4 = 4608 combinations Ensemble-based process studies and quantification of uncertainties

Modeled Tskin (July 12th, 21:00 UTC, 2004) Niu et al. (2010b)

Modeled Leaf Area Index (LAI) and Green Vegetation Fraction (GVF) Niu et al. (2010b)

River network model: RAPID Routing Application for Parallel computatIon of Discharge 30

31 Framework for calculation Atmospheric Model or Dataset Vector River Network - High-Performance Computing River Network Model Land Surface Model

River network modeling 32 RAPID Uses mapped rivers Uses high-performance parallel computing Computes everywhere including ungaged locations

Streamflow map 33 Guadalupe River at Victoria Flow calculated everywhere 1) Optimize model parameters 2) Run model

Guadalupe River near Victoria, TX 34

Animation: flow map (Jan-Jun 2004) 35 Thank you to: Adam Kubach, Texas Advanced Computing Center 01/01/2004 – 06/30/2004 every 6 hours (RAPID computes 3-hourly) 5,175 river reaches; 26,000 km 2 River model improves computations over lumped model, compared to observations at 36 gages

Transition of RAPID to other river basins Learning RAPID –Compile and run for an existing case (0-1 months) if basic knowledge of Linux and Fortran Adapting to new domain –Downloading NHDPlus data and preparing input files (2 weeks) –Downloading and processing USGS data (1 week) –Getting gridded runoff information 36

Water chemistry measurements and statistical modeling of nutrient outflow 37

San Antonio, Guadalupe, Mission and Aransas Rivers Watersheds

High resolution sampling Sampling targeted to high flow events that potentially carry more nutrients 39

Measured concentrations

concentration-runoff relationships

DON: concentration-runoff relationships

LOADEST Regression models MODEL 2 Ln(Conc) = a0 + a1 LnQ + a2 LnQ^2 where: Conc = constituent concentration LnQ = Ln(Q) - center of Ln(Q) MODEL 6 Ln(Conc) = a0 + a1 LnQ + a2 LnQ^2 + a3 Sin(2 pi dtime) + a4 Cos(2 pi dtime) where: Conc = constituent concentration LnQ = Ln(Q) - center of Ln(Q) dtime = decimal time - center of decimal time

Model fit: R 2 values (expressed as %)

Model output: concentration time series

Model output: flux time series

Summary Collected and analyzed 2-yr sampling data in Guadalupe, San Antonio, Mission, and Aransas rivers. Noted that general patterns of DIN dilution and DON enrichment during storm flows are similar across all four rivers. Major differences in nitrate concentrations among rivers (during low flows) reflect differences in anthropogenic nitrogen sources. 47

Estuary model 48

Study Area Two River Basins Guadalupe San Antonio Four HUCs in each basin Guadalupe Estuary Centrally located along Texas coast Microtidal Small bay area but large watershed relative to other Texas systems

A generic ecosystem model (3 components with 2 boundary conditions) Mass-balance model Two boundaries: LGRW & LSRW Three components: Nutrient (DIN) –Phytoplankton – Zooplankton Re-mineralization and implicit sinking (or horizontal exchange) were assumed to be 50%, respectively Δ=1 hr & RK 4 th order scheme

Model Results No loadings (both boundaries shut down): No loadings (both boundaries shut down): Initial nitrogen pool for DIN, Phyto and Zoo will get eventually depleted When LSRW (2 nd panel) or LGRW (3 rd panel) were open: discharged DIN kept nitrogen pool for DIN, Phyto and Zoo to a certain level LSRW and LGRW had a different timing, duration and magnitude in responses of DIN, Phyto and Zoo Source: Arismendez et al. (2009) Ecol. Informatics 4:

Model Conclusions and Discussion Estuary response differs with respect to varying nutrient concentrations. Lower San Antonio River is delivering more nutrients and driving greater ranges of ecological response than the Lower Guadalupe River. Increases in nutrient concentrations due to human alterations of the landscape may result in future eutrophic conditions in the Guadalupe Estuary.

Conclusions 53

Main thoughts Coastal processes are truly and highly interdisciplinary. Collaboration requires careful planning, frequent communication, and high-level patience. Training next-generation students and postdocs are rewarding and challenging. 54

Accomplishments Peer-reviewed Publications –10 published –10 in press, submitted, or in preparation 2 Ph.D. Dissertations and 2 M.S. Theses; 4 Post-docs Numerous conferences and Proceedings An Integrated Framework of Models and Datasets –Noah-MP for climate and hydrology studies –RAPID river flow model –Water chemistry data Organized 2 nd NCEP/NOAA Workshop of Numerical Weather and Climate Modeling, Austin, Texas, April 19-21, 2010 –Next-generation Land Surface Modeling; transition research to operation –Beyond Land-Atmosphere Interactions River flow and nutrient exportscoastal processes 55

56 New IDS Zong-Liang Yang (UT Geological Sciences) Climate modeling Land surface modeling Hongjie Xie (UTSA Geological Sciences) Remote sensing analysis LULC, NEXRAD Wei Min Hao (U.S. Forest Service) NASA satellite land datasets David Maidment (UT Civil Engineering) Hydrology, river routing Land/ocean coupling, BGC cycles Collaborators: NCAR, USGS, … Jim McClelland (UT Marine Science) Paul Montagna (TAMU Corpus Christi) Estuary ecology Coastal BGC

Study Domain Community-based sampling program: recruiting/training volunteers living near downstream locations; primarily focus on storm-flow conditions, also collect quarterly samples during baseflow to compare with TCEQ or USGS.

Research Questions What are the entire pathways of particulates and solutes from the atmosphere through terrestrial and riverine environments to the coastal waters in the west Gulf of Mexico? How are these pathways affected by climate change and land use change? What are the effects of atmosphere dry and wet nitrogen deposition on riverine nitrogen exports and estuarine ecosystem functions? What are the effects of changing climate and land use on terrestrial runoff and associated nutrient export to and availability within ocean margin waters? 58

59 Thank you! Liang Yang (512)