Summary: Course Literature

Read the summary and the most important questions on Course literature

• 1 Mod 1: Course info and context: The evolution of modern hydrology (from watershed to continent in 30 years) - Eagleson 1994

• The so-called systems approach is central to much of the work described in the paper by Eagleson. What is meant by systems approach, and can you give 2 examples from the discussed research that use this approach?

  Viewing the (modeling) processes not as separate elements, but as one interactive system with complex feedback loops, highly interlinked processes, etc. 
  
  2 examples: 
  

  • The realization that the hydrologic system was not just a cause-and effect system, but all elements are highly interconnected. This realization came due to the need for international cooperation and colour photographs of Earth. 

  • The effects of El Nino helped with the realization of the interconnectednes of the water system at global scales and the coupling of ocean and atmosphere.

• Which key challenges were identified by Eagleson in 1994, and what can you say about their relevance today?

  • Grid sizes and scales issues, with according simplifications

  • Coupling of land surface-atmosphere feedback loops
  • Extremes forecasting
  • Finding simplified model parameters that are:
    Physically meaningful, yet analytically insightful 
  • Moving from engineer-only perspective to interdisciplinary science 
  • Human influence and how to (de)couple this in modeling 

  
  
  Many of these issues persist today, although progresses have been made. Mainly for gid sizes (also due to increased computational power), improved human feedbacks in modeling, and moving into an interdisciplinary power have improved since then. They do however, remain unfolding. 

• What is meant by the hydrologic scale, and how is this different from the catchment scale?

  The hydrologic scale: the effective lateral shift of the water molecules constituting Pm as they leave the surface in E and return in P.  The hydrologic scale is the effective radius of influence of land surface change, to which ET is highly sensitive.
  
  The catchment scale is defined by flow boundaries and the hydrologic scale thus by lateral water shift. Hydrologic scales could vary largely in size, depending on (local) meteorology/climate and vegetation (e.g. Tropical forests might have a very small hydro scale).
  
  Pm/P hereby gives the recycling ratio.

• Explain what the recylcing ratio is and whom developed this definition. What does it indicate in terms of E/S/P and what can it be used for in modeling?

  Budyko: Pm/P = 1/ (1+2wu/EL) 
  
  This indicates that P is a function of S through E. 
  P = precip, S = soil moisture state, E = ET
  
  The recycling ratio can be used for land surface-atmosphere coupling.

• Rewriting the recycling ratio will result in 2 specific graphs. What are the typical shapes and what do they indicate?

  Bimodal soil moisture state as a result of land surface-atmosphere feedback. This occurs if the Budyko paramter B's variance is large enough.
  
  Pinball bimodal graph, explaining preferred model states in which soil moisture can reside. The steep ridges in the graph indicate
  prolonged residence times followed by abrupt transition to the other modal. 
  
  

• 2 Models and model evaluation

• 2.1 .... - Scaefli & Gupta 2007

  This is a preview. There are 5 more flashcards available for chapter 2.1
  Show more cards here

• Study Eqs 1 and 2 in Schaefli & Gupta and describe their difference.

  Where NSE uses the mean observed Q over the entire simulation period, BE uses the benchmark model discharge at time t. An example of a BE value used could be an ensembled mean/interannual mean value for a calendar day.

• Give the eqs for NSE (Nash Sutcliffe Efficiency) and BE (Benchmark Efficiency). What are they used for?

  NSE: See pic
  
  BE: same formula, but mean Qobs is changed by qb(t), meaning the benchmark model Q at time t. 
  
  Used to quantify relative model performance.

• From the paper by Schaefli & Gupta, it is clear that different evaluation criteria and ways to apply these criteria exist, and that these can impact not only inferred model parameters but also whether a simulation is judged good or not. What can you identify as a central element in model evaluation?

  NSE, BE, APB, ASPB all consider a form of noise-to-signal ratio, thus how far off is the simulated value to the observed/benchmarked value/reference model value? The mean is a very central component. Sometimes is almost seems as if they are matched to each other as a way of calibration. 
  
  

  • Mean
  • Comparison/benchmark models
  • Explanatory power
  • Uncertainty
  • Simplification
  • Relevance of context         

• Formulate in your own words what is Gupta et al mean here: Clearly, however, the vast majority of the ‘up—and down’ hydrologic model response is caused by the driving variables,and what we need to measure is how well the catchmentprocess modification of this behaviour has beenreproduced. 

  That you cannot only judge if your model works well by looking if it represents the magnitude of the peaks at the right timing, but the point is to how well the model simulates the processes that cause this result.

• What is the idea of the APB model?

  
  The idea is that this benchmark ‘model’ projects the frequency variation of the driving variables into the output while having
  the correct runoff ratio.