Uncertainty analysis on the parameterization of processes of the biosphere and hydrosphere at the earth's surface in atmospheric models


Principal investigator: Nicole Mölders

Graduate student: Mihailo Jankov

Undergraduate students: Leslie Prochaska, Amy Rulo


-Brief summary of activities and achievements

The processes of the soil and the vegetation are the only natural boundary conditions in Chemistry-Transport-Models, models of the planetary boundary layer, atmospheric and climate models. Since the accuracy to which the exchanges of energy, water, and matter can be determined at the heterogeneous interface biosphere-atmosphere strongly determines the reliability of the afore-mentioned models, it is desirable to know the uncertainty that is caused by the necessity to parameterize and to chose parameters. Such uncertainty analysis was performed for various parameterizations used in state-of-the-art land surface models (LSMs). In doing so, prognostic equations to determine the error due to the parameters were derived and implemented in MM5 that was used as a test-platform. These equations permit - similar to data obtained by experiments - to give an error of the fluxes and quantities simulated at the interface biosphere-atmosphere. Furthermore, the applied modules were evaluated by means of already existing data. The activities allowed to assess which parameterization requires which accuracy of the parameter used. The studies helped (1) to minimize prediction errors, (2) to improve the results of LSMs, and (3) to increase the confidence in the results of studies on land-use changes.


-Results

The project contributed to the knowledge on the impact of feedback processes at the interface biosphere-atmosphere. The sensitivity of fluxes simulated at the interface biosphere-atmosphere were analyzed with respect to the parameterizations applied. The project, moreover, contributed to the understanding of the interactions between the cycles of energy, water, and matter. The investigations improved the understanding of modeling the earth-system. The project is related to IGAC and BALTEX.

The main differences of the actual model concepts of the different LSMs were analyzed. The models were categorized in various classes. Out of these classes representatives were chosen with which further investigations are carried out. SURF represents modules of the force-restore-type like they are used, for instance, in GESIMA, METRAS, FITNAH, REMO, GME, DM, and LM. HTSVS stands for the class of modules using coupled energy- and water-budget equations as well as coupled equations of the soil moisture and soil temperature like they are used in MM5, GESIMA (Leipzig), MEPHISTO etc. OSULSM is chosen to represent the models that work with a skin temperature and decoupled equations for soil moisture and soil temperature as applied, for instance, in MM5, EURAD, or KAMM-Drais. For the soil-vegetation modules, SURF, HTSVS, and OSULSM, stand-alone versions were built that can be driven by observations (temperature, wind, moisture, pressure, etc.). HTSVS and SURF were evaluated by data sets gained in Leipzig. HTSVS and OSULSM were evaluated by data (e.g., fluxes of sensible and latent heat, soil heat fluxes, soil moisture, soil temperature) of the CASES97-experiment. Moreover, HTSVS was evaluated with a more than 5-year data set. By applying the stand-alone versions of the three modules 1D-simulations were performed for examining the often strongly variable plant-physiological and soil-physical parameters. The sensitive parameters were determined. For HTSVS, it was started to derive the equations for the sensitive parameters by differentiating the flux equations with respect to these parameters (uncertainty analysis of first order).

Further results are described in the reports and publications listed below.

Annual report for 2001

Annual report for 2002

Progress report on uncertainty analysis of quantities predicted by OSU land surface model (Prochaska and Mölders 2002)

Progress report on uncertainty analysis of quantities predicted by the hydro-thermodynamic soil vegetation scheme (HTSVS) (Rulo and Mölders 2002)

Progress report on uncertainty analysis of quantities predicted by OSULSM (Jankov et al. 2003)

Progress report on evaluation of parameterizations within MM5 (Cherry and Mölders 2003)

Annual report for 2003 (in German); (in English)

Final report (in German)

Jankov, M., 2005: Analysis of uncertainty in simulated exchange of heat and moisture at the land-atmosphere interface, MS thesis


-Peer-reviewed publications resulting from the project

Mölders, N., Jankov, M., Kramm, G. 2005. Application of Gaussian error propagation principles for theoretical assessment of model uncertainty in simulated soil processes caused by thermal and hydraulic parameters. J. Hydrometeorol . (in press)

Mölders, N., 2005. Plant and soil parameter caused uncertainty of predicted surface fluxes. Mon. Wea. Rev. (in press)

Mölders, N., Olson, M.A., 2004. Impact of urban effects on precipitation in high-latitudes. J. Hydrometeor. 5, 409- 429.

Mölders, N., Haferkorn, U., Döring, J., Kramm, G., 2003. Long-term numerical investigations on the water budget quantities predicted by the hydro-thermodynamic soil vegetation scheme (HTSVS) – Part II: Evaluation, sensitivity, and uncertainty. Meteorol. Atmos. Phys. 84: 137-156.


- Acknowledgement

This research was performed in cooperation with University of Leipzig. It was finanially supported by BMBF under contract 07ATF30. Cooperation existed with various groups from DEKLIM. For the evaluation studies data from ATLAS and WINTEX were used. We thank the ATLAS and WINTEX people for providing the data. We also thank those scientists who sent us descriptions of their models.


For further information contact molders@gi.alaska.edu