Palmer Long-Term Ecological Research on the Antarctic Marine Ecosystem: An Ice-dominated Environment Physical Oceanography Component: CTD, Modeling and Analysis 48 months, requested starting date 1OCT98 YEAR 1: 10/1/98-9/30/99 YEAR 2: 10/1/99-9/30/00 YEAR 3: 10/1/00-9/30/01 YEAR 4: 10/1/01-9/30/02 Douglas G. Martinson Summary This proposal requests funds to add Dr. Douglas Martinson as a Co-PI to the Palmer LTER (PAL) to provide the physical oceanography (PO) component of the LTER PAL proposal. The key objective of my participation is to provide physical oceanographic temperature and salinity profiles (CTD measurements) critical to the interpretation of the ecological system and its relationship to its physical setting. I will also provide a modeling and analysis basis from which the physical-bio-geochemical system can ultimately be simulated, synthesized, analyzed and interpreted. 1.0 Introduction This proposal represents the basic physical oceanography (PO) component of the LTER Palmer Long-Term Ecological Research (PAL) program as articulated in Smith et al. (1996). Specifically, this component provides hydrographic and modeling support. Consequently, the scientific details, hypotheses and goals articulated in that LTER proposal are not reiterated here. Rather, I only repeat that information required to provide the broader context of the currently proposed work. 1.1 Working Hypothesis The PAL program is focused on: "understanding the ecological role of sea ice with the primary object being to gain a general understanding of the physical and climatic controls on interannual sea ice variability, the effects of this variability on trophic interactions, and the biogeochemical consequences thereof". The PAL observational and experimental programs reflect this primary research goal. The central null hypothesis ("Neither the presence nor the extent of annual sea ice in the PAL study area influences ecosystem structure and dynamics."), and several alternate hypotheses are discussed at length in the original proposal. The alternate hypotheses describe varying degrees to which the sea ice may influence the ecological system (invalidating the null hypothesis). 1.2 Purpose As the sea ice distribution is intimately coupled to the evolution and structure of the underlying ocean, the PAL program objective demands a comprehensive sampling strategy involving ecological as well as physical variables. It also demands a modeling component with which the complex multi-disciplinary observations can be synthesized, hypotheses tested, system sensitivities assessed and sampling strategies refined. It is the purpose of this PO subcomponent to satisfy these needs (in place of the previous physical oceanographic/modeling group which has withdrawn from the program) by specifically providing: requisite physical oceanographic observations and analyses, and modeling capabilities. 1.3 Objectives The particular objectives of this program, are twofold: (1) Oversee the shipboard collection of ocean temperature and salinity (CTD) profiles; process these observations and distribute the processed profiles to the LTER community and relevant data centers. (2) Make available and employ a continually evolving hierarchy of models that will prove useful in evaluating and analyzing the PAL study area physical-bio-geochemical system evolution; introduce model subcomponents and other necessary modifications required to refine our modeling capabilities; perform relevant experiments; and, determine, when possible, local scaling relationships, bulk parameters and other system reductions that aid in our observation analysis, synthesis and interpretation. 2.0 Approach Limited resources preclude both a dedicated physical oceanographic CTD/hydrographic field program and extensive analysis/modeling study. However, we believe that the program objectives can be achieved with a relatively moderate multi-PI/ASA cooperative CTD field effort coupled with an incremental modeling/analysis effort periodically supplemented through externally funded complementary studies. 2.1 CTD Observations It is possible, with a team effort and clearly defined responsibilities, to eliminate the otherwise costly PO-dedicated field contingent. This reflects the fact that Seabird CTDs provide such good accuracy that they obviate the need for calibrating the salinity against a salinometer in a program of this nature (though this is not true for programs requiring WOCE-quality CTD data). The quality control, heretofore provided by the shipboard salinometers, can now be obtained by inter-sensor calibration using the shipboard Seabird CTD that has been recently modified to include double temperature and conductivity sensors (ASA will calibrate them before and after each cruise as part of their regular maintenance of the CTD equipment). With the calibrated double sensor configuration, the physical oceanographic field component need not be burdened with the labor-intensive bottle sampling hydrography and lab work. This allows a relatively manageable field effort since the shipboard operations reduce to fairly standard and automated tasks. (ASA, which owns and maintains the CTD, provides the shipboard maintenance of the equipment.) Consequently, the field party will not include a PO-dedicated specialist, but rather, the CTD will be operated at sea by the (non-PO) members of the LTER field-going scientific party that will be overseeing the Niskin bottle sampling component. To assure high quality, I will put in place a standard CTD operating procedure for the field party to follow (I will also participate on the 1999 winter cruise to help refine this procedure, but that is discussed and budgeted under a separate supplemental proposal). The operational guide will include instructions for operating the CTD. It will also include the email transmission of newly collected CTD data each day back to Lamont where we will provide quality control, early diagnosis of potential problems and identify potentially interesting scientific features that may warrant further attention. The data will be delivered to me following the cruise, along with the pre- and post-cruise calibration constants for the sensors. I will run it through our standard processing routine and enter it into the LTER data base as soon as the processing is completed (preliminary data will be available immediately). I will also participate in designing or refining sampling strategies, in addition to the operational needs and details of the shipboard CTD operation prior to the cruises. This operational strategy has yet to be tested on a regular multi-year basis. Our experience to date, including our highly successful one-year CTD operation in the Arctic during the SHEBA (Surface HEat Balance of the Arctic) experiment, suggests that it should work well. However, as the CTD data are crucial to the PAL field program, if we find that this strategy can not be relied upon to provide consistently high-quality CTD data, the PAL PIs will reassess the program and make the necessary shifts in funding to establish an operational procedure that will satisfy this need (likely by including a dedicated CTD expert in the field party each cruise). 2.2 Modeling/Analysis The modeling/analysis will be coordinated to make efficient use of resources. In particular, every effort will be made to use my existing hierarchy of models and standard analysis techniques, with minimal model development within this program, though some will be required to include the additional biogeochemical variables. Analysis Initially, we intend to apply a set of analyses, designed and tested for Weddell gyre CTD data, to the existing PAL data base. The methodology involves vertical integration of the temperature and salinity data which provides a set of physically-meaningful bulk property distributions. These distributions provide information regarding the mean seasonal exchange of deep and surface waters, bulk stability of the water column (and its susceptibility to overturn and ventilation) and in situ sea ice formation constraints. The methodology and analysis insights are described in detail in Martinson and Iannuzzi (1998), and not repeated here. With the PAL data, this analysis should provide a robust set of parameters for displaying the temporal and spatial distributions of the physically-meaningful water column characteristics (this will help refine our sampling strategies in future years). This in turn, when compared to analogous bulk property distributions of the biogeochemical properties should help establish the covarying nature of the physical and biogeochemical components of the PAL region. These analyses, including the development of the appropriate biogeochemical bulk properties and establishing their covariation with the physical parameters, should occupy the first two years of this program. Modeling Upon establishing the covarying components of the system, we then intend to employ the various models currently available in order to investigate the mechanisms, interactions and feedbacks responsible for the observed covariations. Currently, I have the following modeling capabilities: (1) a 1-dimensional, in the vertical, upper ocean - sea ice model, to which a basic bio-geochemical model is currently being coupled under a separate grant; (2) two different 3-dimensional regional primitive equation models (based on the Rutgers, S-Coordinate Rutgers University Model, SCRUM; and the Miami Isopycnal Coordinate Ocean Model, MICOM) coupled with sea ice models; (3) the Goddard Institute of Space Studies (GISS) global ocean-atmosphere coupled general circulation model (GCM) including sea-ice; and, (4) two different versions of the GISS atmospheric GCM coupled with my one-dimensional upper ocean - sea ice model. I am also funded externally to: (1) add improved dynamic and thermodynamic sea ice modeling formulations to all of the modeling capabilities listed above; (2) perform cross-scale evaluations and inter-model comparisons, testing and improving the degree to which the upper ocean - sea ice processes are represented in the large-scale models, and evaluate interactions and feedbacks across the different scales; and (3) continue to improve the model upper ocean - sea ice representation in the GISS GCMs. The specific model employed to investigate the observed relationships will depend upon the developments and improvements that transpire over the next couple of years (that is, during the time we are focused on the observational analyses). However, in general, we intend to begin by using that model (or those models) that appear to be most appropriate for addressing the relationships that we uncover. The model study will focus on performing a number of experiments designed to evaluate the ability to simulate not only the observed distributions focusing on the climatology (local and regionally-averaged), and statistical and specific variability, but in capturing the more difficult, but meaningful, covarying relationships of the physical and biogeochemical bulk properties. These represent a far more rigorous test of the model capabilities than simple comparison to single variable profiles. The degrees of success in these initial simulations will dictate the specific model emphasis, required model improvements, additional experiments, and data-model comparisons, that will then follow. Ideally, if we are capable of simulating key observed covarying characteristics, we will then perform a set of experiments designed to identify the underlying mechanisms, processes and feedbacks. We will also assess the sensitivities of these covarying relationships to changes in the forcing, initial conditions and changes in any one particular component of the overall physical-biogeochemical system. As the externally-funded programs complete their model-development phases, we will incorporate the improved models into this program, and perform additional experiments, testing the sensitivity of previous results to improved parameterizations and the influence of specifically-added processes. When reasonable simulations are obtained (for specific characteristics of the system, or some aspect of the evolution), I will attempt to determine the external parameter dependence of the system being simulated. If successful, this provides scalings that encapsulate the fundamental internal mechanisms and interactions driving the seasonal, annual and interannual evolution of the coupled system. This allows quantification of the sensitivities of the system to external forcings and changes in local or regional environmental characteristics. It also indicates the algebraic combination of easy-to-observe parameters in forms that capture the essence of the system, unlike a simple variable that may, or may not, clearly represent the fundamental signal contained within the complex noisy system. However, this is an extremely time-consuming and difficult task, and it's accomplishment in this phase of the PAL program cannot be guaranteed - rather it serves as a long-range objective. 3.0 Expected Significance The significance of this program comes through two means: (1) collecting and distributing the fundamental physical observational data base requisite in building our understanding of the relative roles and interactions within the physical-bio-geochemical system; and (2) providing a multi-disciplinary modeling and analysis foundation from which we can build our system understanding while helping to integrate the otherwise disparate data into a common framework. This second contribution ultimately serves to provide a key means through which the various disciplinary elements for the PAL program will be integrated and synthesized to provide a unified and holistic perspective on the system evolution. 4.0 Budget and Budget Justification I will allocate the vast majority of the funding to cover one month of my salary per year and 2.5 months per year for my technician, Richard Iannuzzi. This will allow Rich to perform the processing of the CTD data and the computational aspects of the analysis and modeling work, all of which are fairly computer intensive. I will perform the quality control for the CTD profiles being sent from the cruises to Lamont, and will oversee all of the analysis and modeling work. We have also budgeted for nominal travel (to meet with the LTER group in Santa Barbara once per year), and some nominal lab cost (e.g., computer subscription, phone bills, etc.) No funds are requested for ship time.