Hydologic Models

Hydologic Models

Modeling is the process of creating a representation (model) of something usually larger and much more complex. It allows a better understanding of real systems for problem-solving and for analyzing what-if scenarios. A hydrologic model can be a small-scale physical model (like a miniature) or a computer simulation that uses algorithms and equations to imitate real-world watersheds. When water enters a watershed as rainfall, it may undergo several things; it may be intercepted by leaves, stored in lakes, reservoirs, or underground aquifers, flow overland surface, flow through rivers before exiting to the sea, etc. Hydrologic models are used to understand how a watershed responds when subjected to rainfall and to know how water moves within the watershed. Hydrologic modeling has a wide range of applications: water resources planning, flood forecasting, climate change impact assessment, sediment transport, and water quality modeling, among others.

At Project e-SMART, we conduct hydrologic modeling through computer simulations to determine temporally varying discharges that enter Manila Bay from its surrounding watersheds. We can model how the water in Manila Bay moves (through hydrodynamic modeling) and understand how pollutants are entering and moving within the bay.

Developing a model starts with defining what processes are important and identifying model parameters that control these processes. The accuracy of the results of the model analysis depends on how well the model represents reality. That’s why it is important to collect relevant input data. Inputs to hydrologic models can be categorized into topographical and meteorological data. The characterization of the topography of the area being modeled involves the use of terrain, land cover, and soil maps. The meteorological data on the other hand include rainfall, temperature, wind speed, relative humidity, and solar radiation.

Project e-SMART has gathered these data from different concerned agencies. The terrain and land cover files were obtained from (National Mapping and Resource Information Authority) and the soil profiles can be obtained from (Department of Agriculture – Bureau of Soils and Water Management ). The weather data were obtained from (Dost_pagasa), (DOST Advanced Science and Technology Institute), and (Effective Flood Control Operation System). Other sources of data such as global weather (https://globalweather.tamu.edu/) and global soils profile (http://www.fao.org/soils-portal/data-hub/en/) are also accessible online. These coarser datasets can be used in the absence of local datasets in some areas or some periods of time.

After simulating a hydrologic model, the accuracy of its results (model performance) is tested through calibration and validation of the model. These processes involve comparison of simulation results such as flows or discharges with that of actual river discharges. The goodness of fit or nearness of the simulated to the observed can be measured by various statistical parameters.

Calibration is the iterative process of adjusting the model parameters so that the simulation results get as close as possible to the observed data. This is comparable to adjusting the antenna of a television to have clear reception from a selected TV station. Once the model is calibrated well enough, the model is tested for a different time period through validation. Going back to the TV analogy, when an acceptable quality of reception is achieved, it can be expected that other channels will have better reception. Of course, it may not be always the case, as there may be other factors affecting TV reception. Similarly, there may be other reasons why a model can work for a certain simulation period and not work well on another.

It’s also good to note that the quality of calibration and validation also relies on the quality and quantity of observed data from the water level or discharge gauges.

The main output of a hydrologic model is the discharge at the outlet of the watershed. The flows can then be assessed to know certain statistics such as daily, monthly, or annual averages. The average flows during dry and wet seasons, or seasonal means, can also be obtained as shown in the image.

The seasonal means for each river draining to Manila Bay and Laguna de Bay resulting from the hydrologic models can be compared and analyzed. For example, the Pampanga River Basin contributes the most freshwater flows to Manila Bay due to its size (around 79% of all flows during the wet season and around 82% during dry season).

Aside from discharges, other results from hydrologic models include calibration plots which measure the accuracy of the models, and water balance distributions which describe the hydrological response (how much water flows overland, how much seeps through the ground, etc.).