Earth Science Department SeminarExport this event to calendar

Friday, February 8, 2019 — 2:00 PM EST

Challenges and Opportunities in Integrated Hydrosystem Modelling: A Divide-Reduce-and-Conquer Approach to Overcome the Curse of Dimensionality

Presenter
Dr. Young-Jin Park
Earth and Environmental Sciences
University of Waterloo

Astract
Integration of multiple hydrologic, geochemical, and ecological processes into a single analysis framework is destined to increase the dimensionality of the problems to solve. To overcome the curse of dimensionality in integrated hydrosystem modelling (IHSM), a divide-and-conquer approach is discussed as a potential solution strategy. In the approach, the complex models are divided into the subproblems, the solutions of which can be obtained independently, yet interacting with their neighbours. It is demonstrated that the efficiency and robustness of the approach depends on the sizes of the intersections among the sub-problems. The results suggest that the optimal design of sub-problems is a key for success in its applications. The scale is an important dimension to be solved in IHSM. To simultaneously solve those processes of different scales, a divide-reduce-andconquer approach is discussed. A necessary and sufficient condition is derived to apply the approach when solving two interacting sub-problems of different scales.

Location 
PHY - Physics
145
200 University Avenue West

Waterloo, ON N2L 3G1
Canada

S M T W T F S
27
28
29
30
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
  1. 2019 (39)
    1. December (3)
    2. November (2)
    3. October (1)
    4. September (2)
    5. August (1)
    6. July (2)
    7. June (2)
    8. May (3)
    9. April (4)
    10. March (4)
    11. February (8)
    12. January (7)
  2. 2018 (45)
    1. December (2)
    2. November (6)
    3. October (4)
    4. September (7)
    5. August (3)
    6. July (3)
    7. June (5)
    8. May (2)
    9. April (2)
    10. March (7)
    11. February (2)
    12. January (2)
  3. 2017 (9)
  4. 2016 (9)
  5. 2015 (15)
  6. 2014 (15)
  7. 2013 (20)
  8. 2012 (8)