CVEN9612 – Catchment Modelling
Assignment 1 Part 2 – SMART Modelling (10 marks)
This question is an extension to the SMART tutorial over a sub-basin of First Creek at Waterfall Gully (A5040517) that you did in the computer lab workshop. For this assignment, it is required for you to process additional simulations for the whole basin as outlined in Questions 1) and 2), with following information:
- Study period: 01/01/1984 – 31/12/1984 (1 year, to be specified in “UserRunInfo.m”)
- 1-year climate input file: “Input_Climate_1year.csv” in “Toolbox_Input” folder
- “extract_simulation.m”: a MATLAB script. to extract simulated results
- “Observation.xlsx”: observed flowrate (Q), surface soil moisture (SM) and evapotranspiration (E) to be compared with the simulations
1) Process two simulations with "DISTRIBUTED PIXEL" and "ECS Left bank/right bank/headwater (3ECS)”
Then compare and discuss the results in terms of 1) computational time and 2) model performance compared
to the observations given in “Observation.xlsx”. The model performance should be assessed by calculating
three statistical metrics (i.e. mean square error, Nash and Sutcliff efficiency and Pearson’s
correlation) between the simulations and observations. (Please refer to the uploaded journal paper
“Evaluating the use of “goodness‐of‐fit” Measures in hydrologic and hydroclimatic model validation” for the three statistical metrics) (5 marks)
2) The goal of this question is to assess the impact of land cover change on simulated fluxes in the catchment. To do this, use the new land cover map (“crop.tif”) that is provided in the
“SMART\Toolbox_Input” folder and only implement the 3 ECSs delineation to assess the impact of land cover change in the whole basin. Discuss the differences in water balance components between this
simulation and the simulation with “trees.tif” data. Which of the water balance components show the largest change with your first simulation? Explain the main factor causing the differences in the
simulated results? (5 marks)
