Jia Wang

Introduction

Jia Wang obtained his Ph.D. in atmospheric and oceanic sciences in 1993 at McGill University, Canada. He worked at the Bedford Institute of Oceanography and the University of Miami before joining IARC-Frontier as associate professor and group sub-leader in January 1998. His research interests include Arctic climate change study, coupled ice-ocean modeling, and physical-ecosystem modeling.

Research

1) Arctic Climate Change Studies: Following the finding of the Arctic Oscillation (AO), I proposed the corresponding Arctic sea ice oscillation (ASIO, Wang and Ikeda 2000, 2001). I continue to conduct studies of Arctic climate, involving interaction/feedback among atmosphere, ocean, sea ice, and the freshwater budget (terrestrial process). Recently, I proposed that the north hemisphere sea ice reduction trend in the last three decades triggered the quasi-decadal sea-ice oscillations. This hypothesis has been confirmed using both conceptual models and data analysis. Further testing using coupled ice-ocean general circulation models (GCMs) and coupled atmosphere-ice-ocean GCMs is under way. I also found that during the last three decades, Arctic freshwater storage decreased, as sea ice extent shrinked, from intra-seasonal to inter-decadal time scales (Figure 1).

2) Development of Arctic Coupled Ice-Ocean Models: I am developing regional (nested Beaufort Sea) and pan Arctic-North Atlantic Ocean coupled ice-ocean models to study the interaction between ocean circulation and sea ice, and dense water formation in high-latitude oceans. I conducted research projects on numerical simulations of the ice-ocean system in the pan Arctic and North Atlantic Ocean (Figure 2). I used the Princeton Ocean Model and a full dynamic- thermodynamic ice model with seven categories. I introduced new rheology for sea ice dynamics based on the theory of elastic-plastic-viscous flows. This coupled model is being used to study sea- ice and ocean circulation anomalies and climate variability in the pan Arctic-North Atlantic. The nested Beaufort Sea (regional) ice-ocean-oilspill model (Figure 3) will be used to perform operational prediction with the help of assimilating observed and remote sensed data for the coastal oil exploration. I am also developing a coupled ice-ocean model (MOM) to study pan Arctic-North Atlantic-North Pacific ice-ocean climate in response to atmospheric forcing in a T3E supercomputer.

3) Development of Coupled Biological-Physical Model: I am developing a 3-D coupled biological-physical model. The biological model consists of nine compartments: nutrients, phytoplankton, zooplankton, and detritus. The model has been applied to Prince William Sound, Alaska, and is being applied to the Bering-Chukchi seas (Figure 4). This model will be further coupled with marine carbon cycle and other biochemical processes in the region. The new model includes nitrite-nitrate, ammonium, silicon, two types of phytoplankton and three kinds of zooplankton, involving an ocean DMS (dimethyl sulfide) component. The upper mixed layer mixing is the key dynamical process to the DMS cycle/pathway.

4) Development of Coupled Hydrological-Ocean Circulation Model: I am developing a coupled hydrological-circulation model in the Gulf of Alaska (Figure 5). Freshwater runoff of the line source origin was found to be nearly twice as large as the point source (rivers). This hydrological (DEM) model will be coupled to the ocean circulation model (Figure 6). This coupled model has a potential to study freshwater fluxes, DOC budget and biogenic matters from land to the ocean. This coupled hydrological-ocean model will be further coupled to a biological model in the Gulf of Alaska.

5) Development of statistical methods/analyses to study Arctic climate change, such as EOF, SVD, complex EOF, and wavelet analyses. These methods have been used for analyzing massive datasets in oceanography and meteorology in the Arctic (Figure 7).

6) Development of the data assimilation methods for ice, ocean 4-D data assimilation, such as the direct nudging method, sequential theories (optimal interpolation, Kalman filter, and extended Kalman filters), and variational methods (adjount method). The direct nudging method has been used to predict ocean circulation and thermohaline structure in a nowcast/forecast system (Figure 8).

(click on images for larger view)
Figure 1-1: (Upper panel) The numerical solution of the Arctic and Nordic Seas upper-layer thicknesses, H1 and H2 , responding to an increase in w (i.e., decrease in sea ice thickness from 2m to 1.3m) from 6.5x10⁴ m² /s to 9x10⁴ m²/s (t=0 to 87.5 years) and the w is fixed to 100 years (87.5≤to≤100 years). The phase difference is 3π/4, with the Nordic Seas leading the Arctic Ocean. (Lower panel): The time series of sea-ice area anomalies for the Arctic Ocean (blue) and the Nordic Seas (green).
Figure 1-2: Freshwater thickness anomalies in both the Arctic Basin (green) and the Nordic Seas (purple) in both summer (upper panel) and winter (lower). In most cases, the two regions are out of phase, indicating an oscillation is in work, as Arctic sea-ice oscillation (ASIO)
Figure 1-3: Proposed new feedback loop for the observed decadal Arctic climate cycle (Mysak and Venegas (1998) and the observed long-term downward trend due to a positive feedback of ice and clouds. An arrow with a plus sign between box A and box B means that a positive (negative) anomaly in A would cause a positive (negative) anomaly in B after a certain delay, while an arrow with a minus sign would results in a negative (positive) anomaly in B.

(click on images for larger view)

Figure 1-1: (Upper panel) The numerical solution of the Arctic and Nordic Seas upper-layer thicknesses, H1 and H2 , responding to an increase in w (i.e., decrease in sea ice thickness from 2m to 1.3m) from 6.5x10⁴ m² /s to 9x10⁴ m²/s (t=0 to 87.5 years) and the w is fixed to 100 years (87.5≤to≤100 years). The phase difference is 3π/4, with the Nordic Seas leading the Arctic Ocean. (Lower panel): The time series of sea-ice area anomalies for the Arctic Ocean (blue) and the Nordic Seas (green).

Figure 1-2: Freshwater thickness anomalies in both the Arctic Basin (green) and the Nordic Seas (purple) in both summer (upper panel) and winter (lower). In most cases, the two regions are out of phase, indicating an oscillation is in work, as Arctic sea-ice oscillation (ASIO)

Figure 1-3: Proposed new feedback loop for the observed decadal Arctic climate cycle (Mysak and Venegas (1998) and the observed long-term downward trend due to a positive feedback of ice and clouds. An arrow with a plus sign between box A and box B means that a positive (negative) anomaly in A would cause a positive (negative) anomaly in B after a certain delay, while an arrow with a minus sign would results in a negative (positive) anomaly in B.

Figure 2 (click on images for larger view)
Figure 2-1: The pan Arctic and North Atlantic Ocean (PANAO) coupled ice-ocean model (CIOM) domain, topography, and the Hamilton, Fram Strait, and trans-Arctic-Atlantic sections. Depths are in meters.
Figure 2-2: The simulated Arctic sea ice flow patterns in March (upper left) and September (upper right). To show the Arctic sea ice pattern, the velocity larger than 2 cm/s is represented by arrows in red. Shown are also the model-simulated Arctic sea-ice concentration distribution in March (lower left) and in September (lower right).
Figure 2-3: The model-simulated Arctic sea-ice thickness distribution in March (upper left) and in September (upper right). The Arctic sea ice distributions based on limited submarine measurement are also shown for winter (lower left) and summer (lower right) (after Bourke and Garret 1987).
Fig. 2-4: Model-data comparison of the seasonal cycle of sea ice areas in subregions 2, 3, 4, 5, 6, 7 (referred to Fig. 5), and in the whole Arctic region.
Fig. 2-5: Model simulated 500-m (Atlantic Water Layer) circulation (upper panels), salinity (middle), and temperature (lower) fields of March (left column) and September (right).
Figure 2-6: The model-simulated trans-Arctic-Atlantic Ocean T (a) and S (b) sections (see Fig. 1-1 for the location).

Figure 3 (click on images for larger view)
Figure 3-1: A nested coupled ice-ocean model in the nearshore Beaufort and Chukchi Seas.
Figure 3-2: The comparison between the ADCP measurement (Shimada 2000) and the nested model.
Figure 3-3: The comparison between the observed winter halocline ventilation (Melling 1993) and the model simulation.
Figure 3-4: Historical observations derived from WOA01 in the Beaufort Sea: (top) summer surface temperature, and (bottom) summer surface salinity.

Figure 4 (click on images for larger view)
Figure 4-1: Comparison of observed (dashed) and calculated (solid) Chlorophyll concentration at depth of (a) 0m, (b) 5m, (c) 10m, and (d) the mean of 25-50m at AFK station of Prince William Sound, Alaska in spring of 1996
Figure 4-2: Model-simulated time-depth chlorophyll (mg/m³ ) time series (upper panel) and Nitrate + Nitrite (μM) time series (1996) at station AFK.

Figure 5 (click on images for larger view)
Figure 5-1: The Gulf of Alaska model domain for the MITgcm (z-coor) and Princeton Ocean Model (POM, sigma-coor.). The resolution is 4km.
Figure 5-2: Model-simulated surface current in March (upper) and September (lower)
Figure 5-3: Model-simulated surface salinity in March (upper) and September (lower). The summer freshening is captured using our hydrological digital elevation model (DEM) input.
Figure 5-4: Historical observations derived from WOA01 in the Gulf of Alaska: (top) summer surface temperature, and (bottom) summer surface salinity. Data from three other seasons are also processed.

Figure 6 (click on images for larger view)
Figure 6-1. Elevation in the model for the Gulf of Alaska (top) and the watershed of the line source (shaded in colors) and five big rivers (in colors, bottom), which are named (from left to right) Susitna, Copper, Alsek, Taku, and Stikine rivers. The "+" signs denote the NCDC stations available in the study region.
Figure 6-2: Model simulated monthly climatology (seasonal cycle) of freshwater discharge derived from the 41-year simulations (1958-1998) and standard deviations (vertical bars). Black line is the total discharge into the Gulf of Alaska; the blue line is the line source and the red is the point source.
Figure 6-3: Time series of daily mean (top), monthly mean (middle), and annual mean (bottom) discharges of Susitna River from 1975 to 1993. Dashed lines are modeled results, while solid lines are observed data of USGS Susitna River station at Susitna Station.

Figure 7 (click on images for larger view)
Figure 7-1: The time series of first EOF mode for the surface air temperature (upper left) and sea-level pressure (lower left) using data from north of 45°N and their power spectrum of the wavelet transform of SAT (upper right) and SLP (lower right). Decadal time scales are striking.

Figure 8 (click on images for larger view)
Figure 8-1: A Straits of Forida Nowcast/Forecast System (SFNFS) was set up in 1995 (Wang 2001, J Atmos. Oceanic Technol.) using POM.
Figure 8-2: The automatic SFNFS, its flow chart and C-scripts that controls and automatically execute the SFNFS (Wang 1999, IARC/UAF report).

Figure 8 (click on images for larger view)

Figure 8-1: A Straits of Forida Nowcast/Forecast System (SFNFS) was set up in 1995 (Wang 2001, J Atmos. Oceanic Technol.) using POM.

Figure 8-2: The automatic SFNFS, its flow chart and C-scripts that controls and automatically execute the SFNFS (Wang 1999, IARC/UAF report).