Real Time Streamflow Forecasting and Reservoir Operation System for Krishna and Bhima River Basins in Maharashtra

RTSF & ROS VERSION 1

USER GUIDE DRAFT

June 2012

RTSF & ROS USER GUIDE (DRAFT)

Client

Client’s representative

Project

Project No

Authors

Date

17/06/2012 Gregers Jørgensen – DHI Guna Paudyal - DHI 0

Revision

Draft Manual

Description

Key words

Approved by

GHJ

By

Checked

Approved

Date

Classification

Open Internal Proprietary

Distribution

No of copies

Contents 1

INTRODUCTION........................................................................................................... 1

2 2.1 2.2

RTSF & ROS APPLICATION ........................................................................................ 2 Starting RTSF & ROS ................................................................................................... 2 RTSF & ROS User Interface ......................................................................................... 2

3

CONFIGURATION OF RTSF & ROS ............................................................................ 5

4

RTSF & ROS BATCH JOB ............................................................................................ 9

5

RTSF & ROS MIKE11 TOOLS .................................................................................... 11

6

RTSF & ROS WEB PAGE ........................................................................................... 12

7

RESERVOIR OPERATION MODULE ......................................................................... 17

8

SIMULATION OF SCENARIOS .................................................................................. 18

9

DISSEMINATION OF RESULTS ................................................................................. 21

1

INTRODUCTION RTSF & ROS is a decision support system used to assist in the daily operation of a real time forecasting based on MIKE modelling software. RTSF & ROS has been developed as a user friendly system, which works as a stand-alone Windows application. It does not require in depth knowledge of model development and setup and GIS applications. However, knowledge of hydrological, hydraulic and forecast modelling is essential for correct interpretation and dissemination of results. Based on a very simple User Interface it is possible to have full control on the Online Forecasting, which may run automatically. RTSF & ROS also provides a provision to fast run of different alternative scenarios in offline mode and look into old historical forecasts. Another advantage of the less complicated setup is the provision for easy modification of a setup such as forecasting locations, time steps, warning levels etc. The present Draft User guide or Manual describes the operation and functionality of RTSF & ROS. The Manual is subdivided into following chapters: Chapter 2: RTSF & ROS Application Chapter 3: Configuration of RTSF & ROS Chapter 4: RTSF & ROS Batch Job Chapter 5: MIKE11 Tools Chapter 6: WEB page Chapter 7: Reservoir Operation Chapter 8: Simulation of Scenarios Chapter 9: RTSF & ROS Tools The present version of RTSF & ROS works without the real time telemetry data from RTDAS. The system and the Manual will be updated once the RTDAS data will be available .

Page 1

2

RTSF & ROS APPLICATION 2.1

Starting RTSF & ROS RTSF & ROS starts from a command prompt, with the following line input: C:\Floodwatch\Bin\ForecastReaderNew.exe setup.txt Where setup.txt is the input file to the RTSF & ROS application.

The input file “setup.txt” includes following lines: Line 1: Number of Models N lines: Configuration files for each model setup included Last line: Start mode (Online or Offline)

The example below shows the content of a setup file, with 1 model setup included: NumberOfModels="1" Model1_Config="C:\FloodWatch\Models\Krishna\Setup\Krishna_Config.txt" Mode="Offline"

The setup of the model configuration file is described in the next chapter. It is also possible to start RTSF & ROS via a batch job, where the batch job contains the Command prompt line input. Examples of the setup, starting in Online and Off-line mode, are included in the installation folder C:\Floodwatch\setup.

2.2

RTSF & ROS User Interface Figure 2.1 shows the User Interface for RTSF & ROS. The Interface can be used to manage the most common activities in the daily operation of a forecasting system. RTSF & ROS User Interface contains (referring to figure 2.1): A) Tools, selection boxes and status (left column) B) An overview (as bitmap) of the model setup and stations included (upper right) C) Graphical and Tabular View (lower right) Tools, selection boxes and status information Following tools and settings are possible (referring to figure 2.1) 1) Selection of the Actual Model Setup in Online or Offline mode 2) Selection of Time of Forecast 3) Status line 4) Setup – configuration of Model setup in RTSF & ROS (see Chapter 3) 5) RTSF & ROS batch Jobs (see Chapter 4) 6) MIKE11 tools (see Chapter 5) 7) View Flood Map (not yet implemented) 8) WEB page (see Chapter 6) 9) Scenario management (see Chapter 8) 10) Dissemination of Results (see chapter 9)

Page 2

Fig 2.1. RTSF & ROS Interface.

The RTSF & ROS Overview The Overview shows, the river network, the flood status (warning level) for a selected date at forecast locations on the river system (shown as coloured squares at each forecasting location) and the accumulated catchment rainfall (shown as values in grey squares) for a selected period in the modelled catchments.

Graphical and Tabular View Graphs and tables can be selected from the map, when clicking on a station on a map. It is possible to select between water level and discharge from the selection box just above the graph to the left. Zoom facilities are available when right clicking on the map or clicking on the button in the upper left corner. In the upper right corner it is possible to select flood status for different time steps after Time of Forecast. The graphical and tabular view of catchment rainfall can be selected when clicking on a number on the bitmap. The number represents the accumulated rainfall during the period selected in the lower left corner of the graph. The accumulation period represents hours before time of forecast (first selection box) up to hours after time of forecast (second selection box). The tabular view shows the actual rainfall and accumulated rainfall. Similarly, reservoir status can be seen from the reservoir mode selected in the map. The reservoir symbol indicates approximate percentage of fullness of a reservoir. The graphical and tabular views show the reservoir level, inflow and outflow.

Page 3

Online and Offline mode RTSF & ROS can be applied in Online and Offline modes. When running RTSF & ROS in Online mode, the Overviews are automatically updated as soon as a new forecast is ready. Latest Time of Forecast is updated in (2), while the simulation status will appear in the Status box (3). When running in Offline mode it is possible to load historical forecasts and test various scenarios. Scenario simulation is further described in Chapter 8.

Page 4

3

CONFIGURATION OF RTSF & ROS The Configuration Editor is used to prepare or modify a model setup, with all parameters required for the real time operation and presentation of results from the forecasting system. Figure 3.1 shows the Editor used for the RTSF & ROS configuration.

Figure 3.1. The RTSF & ROS Configuration Editor

Referring to Figure 3.1 the configuration editor includes: Model Folder The model folder is the actual location for the model setup. A model folder contains: - MIKE11 subfolder with the MIKE11 model setup - Time series folder with all online time series required for modelling - Setup folder containing details Config-files and spread sheets - Batch folder containing batch files used for forecasts

Page 5

The location of the model folder is: C:\Floodwatch\Models\XXX, where XXX is the name of the actual model (first line in the configuration file). Title Title of the actual forecast simulation Forecast time specifications: -

Time of Forecast (YYYY-MM-DD HH:MM or now (real time))

-

Simulation Period: Historical Period and Forecasting period (in hours)

Forecasting Locations: Forecast locations and parameters related to the forecasting points are specified in a spread sheet (saved in comma separated text format, which can be opened from the configuration editor). New forecast locations can be included as per requirement adding new lines to the spread sheet as shown below: Hydro_Stations Part 1) Specification of ID, coordinates, filenames, forecast location in MIKE 11

Column 1: Included (yes/no) Column 2: Station ID (4-6 characters) Column 3: Station name Column 4: River name Column 5: X-coordinate (projected) Column 6: Y-coordinate (projected) Column 7: X-Longitude (Decimal Degree, 4 decimals) Column 8: Y-Latitude (Decimal Degree, 4 decimals) Column 9: Reference Level (Datum level in meter, 2 decimals) Column 10: File location, Observed Water Level Column 11: File location, Observed Discharge Column 12: MIKE11 Branch name Column 13: MIKE11 WL-Chainage (meter – no decimals) Column 14: MIKE11 Q-Chainage (meter – no decimals) Part 2) Specification of warning levels, quality control and confidence level (optional)

Column 15-18: Warning Levels (Water Levels) as relative level in [cm] (not yet introduced) Column 19-22: Warning Levels (Discharge) in [m3/s] Column 23-25: Quality Control (Discharge) - (not yet introduced) Column 26-28: Quality Control (Water Level) - (not yet introduced)

Specification of colour codes for warning levels: Each colour are represented as a combination of red, green, blue (0-255), see below

Page 6

MQ = Mean Discharge

Meteo stations: The Meteo station spread sheet includes a list of the catchments included on the Overview and on the WEB page.

Column 1: Included (yes/no) Column 2: Catchment ID (4-6 characters) Column 3: Catchment name Column 4: X-coordinate (projected) Column 5: Y-coordinate (projected) Column 6: X-Longitude (Decimal Degree, 4 decimals) Column 7: Y-Latitude (Decimal Degree, 4 decimals) Column 8: File location, Precipitation Note: Catchments need to be ordered as they appear in the NAM parameter file.

Reservoirs:

Column 1: Included (yes/no) Column 2: Station ID (4-6 characters) Column 3: Station name Column 4: River name Column 5: X-coordinate (projected) Column 6: Y-coordinate (projected) Column 7: X-Longitude (Decimal Degree, 4 decimals) Column 8: Y-Latitude (Decimal Degree, 4 decimals) Column 9: Reference Level (absolute level in meter, 2 decimals) Column 10: File location, Observed Water Level Column 11: MIKE11 WL Branch name Column 12: MIKE11 WL Chainage (meter – no decimals) Column 13: File location, Observed inflow Column 14: MIKE11 Q-inflow Branch name Column 15: MIKE11 Q-inflow Chainage (meter – no decimals) Column 16: File location, Observed inflow Column 17: MIKE11 Q-outflow Branch name Column 18: MIKE11 Q-outflow Chainage (meter – no decimals) Column 19: Minimum Water Level of Reservoir (meter, absolute level) Column 20: Maximum Water Level of Reservoir (meter, absolute level) Column 21: Minimum Q (m3/s) Column 22: Maximum Q (m3/s)

Page 7

WEB page: RTSF & ROS includes a post-processing tool used to produce all output for presentation on a predefined WEB page. Two steps are followed in the Web Page functionality. The Web page is first uploaded in a local folder, which can be viewed locally and via an office intra-net. Once it is accepted that the forecasts and other data may be made public, then the web page can be uploaded to the WWW Website. Plots and data will be prepared for all stations included in the Hydro Station, Meteo Station and Reservoir setup files. The Post processing can start from a command prompt or a batch file with following content: “C:\Floodwatch\bin\forecastplot.exe configfile.txt”. Following parameters can be specified for the output to the WEB page: -

Locations of predefined WEB page

-

Data location for the output to the WEB page

-

Time steps included in the Tabular View and in the Flood Bulletin [specified as hours after Time of Forecast]

-

Period for time series plot: Specified in hours before and after time of forecast.

-

Size of time series plot in pixels (Nx: horinsontal size and Ny: vertical size)

-

Time steps included on the WEB page

-

Symbol size in pixels: Size of coloured square of Flood status in the Overview

-

Checkbox for including of observations on the plot

-

Checkbox for including of small time series plot used for output to mobile devices

-

Checkbox for water level adjustment. With this checkbox activated the simulated water level is manually adjusted to the observed water level at time of forecast before forecasts are presented on the WEB page.

-

Checkbox for plot of the simulated water level time series in a cross section. (this option is not activated in the standard version of RTSF & ROS)

Folder Locations: Following folders and files are specified: - MIKE11 simulation file - Location for MIKE Zero bin directory - Mike View setup for a predefined presentation of MIKE11 results - RTSF & ROS Batch file, which can be used to run all forecasting activities - Bitmap used to present data on RTSF & ROS - ArcGIS mxd-file used to open a predefined setup with shape files and Grids. A new bitmap used by RTSF & ROS can be prepared when exporting a map in ArcGIS (save as png-file including write of world file) - Location of RTSF & ROS Archive with zipped historical simulations. The last checkbox is checked when the simulated time series also should be stored in text format.

Page 8

4

RTSF & ROS BATCH JOB The forecasting activities are controlled from a number of Batch jobs combined into one overall batch job, which can run automatic scheduled by the Windows Scheduler or can start from RTSF & ROS. Figure 4.1 shows an example on a combined forecast batch job including a number of batch jobs for each forecasting activity.

Figure 4.1. Example on a combined forecast batch job including a number of batch jobs for each forecasting activity.

Following BATCH jobs are included in the forecasting activities shown in example 4.1. 1) Download of time series data: A automatic procedure which downloads reservoir water level and rainfall data form the Web page of Pune flood control division (www.punefloodcontrol.com). Once the RTDAS is implemented, this functionality will be replaced by import of real time data from the RTDAS data server via the Krishna Bhima Knowledge Base System (KBS) developed and installed at the Control room. 2) Download and & processing of Quantitative Precipitation Forecast (QPF). This is automatically downloaded from the dedicated E-mail ([email protected]), which receives QPF from the RIMES Server automatically every morning. 3) Quality control of data 4) Preparation of simulation file 5) Model simulation 6) Reading of results 7) Preparation of Plots and data for Web page 8) Zipping and archiving of results 9) Saving of forecast in knowledge base

Page 9

10) Completion of forecast The batch job can be checked and changed when clicking on the RTSF & ROS Batch File-Button. The button Start batch job is used to execute the batch job. Status of execution of each batch jobs is monitored via a log file, whose contents will appear in the status field in RTSF & ROS. In the example “Forecast completed” (last line), will be shown in the status field after finishing execution of all batch jobs.

Page 10

5

RTSF & ROS MIKE11 TOOLS RTSF & ROS has following tools which can be used to obtain a more detailed overview of the model simulation: Mike11 Editor: Overview with access to all Model Editors and simulation parameters user for the simulation of the forecast as shown in the figure 5.1.

Figure 5.1. Overview of Model Editors.

MIKE11 logfile: Overview of warning and error messages from Latest forecast simulation: MIKE11 Result Viewer: Detailed Overview of Results from the forecast simulation using MIKE VIEW, with a predefined setup specified via the configuration editor. Figure 5.2 shows an example incl. a river profile, a cross-section and graph.

Figure 5.2. Detailed overview of MIKE11 results from a forecast simulation.

Page 11

6

RTSF & ROS WEB PAGE RTSF & ROS includes post-processing tools used to produce all data and plots required for presentation of the forecasts on a predefined WEB page. The WEB page, based on Google maps includes 4 different data types (Discharge, Water Level, Reservoirs and Precipitation). The post-processing can start from a command prompt as follows: C:\Floodwatch\Bin\Forecastplot.exe config.txt Where the configuration file:”config.txt” includes a complete setup for all stations included on the WEB page. The setup of the configuration file is further described in Chapter 3.

Normally, the preparation of WEB pages takes place in a batch job included as one of the activities for the entire forecasting batch job as described in Chapter 4. RTSF & ROS supports output to two sizes of predefined WEB pages. A large size used for Desktop applications (width larger than 640 pixels) and a small size (width up to 640 pixels) used for Mobile devices (Android / Iphone / Ipad). The Web page can be checked on the local PC from the tool-button: View local Web Page. The uploading of new data with forecasting results to the Internet is controlled from the tool: Export Results to the Internet. Figure 6.1 shows an example on a predefined large WEB page.

Figure 6.1. Example on a Large size RTSF & ROS Web page.

Page 12

Figure 6.2. Example on part of the WEB Bulletin.

In the WEB example on figure 6.1 the following facilities are provided: - Selection of 4 different data types (see selection box below model buttons). In the example data type: Discharge has been selected. - Provision for viewing the forecast in a WEB bulletin (see figure 6.2). - Provision to zoom on the map (zoom bottom to the left of model bottoms) - Provision to select different types of maps (upper left corner, satellite/terrain/roads) - Provision to select different time step on the map (use navigation button left/right on in the lower right corner below the map) - Provision to enlarge the scale on the graph (use zoom button in upper right corner) - Tabular view of forecasted data. Figure 6.3 shows an example on a mobile WEB page. In the example it is possible to select three different data types (Discharge, Reservoir and Precipitation).

Page 13

Figure 6.3.Example on the small Web page for Mobile Devices.

Details of the WEB programming, which is based on open Java script code and Google Maps Api., is not included as part of this manual. The Source code can be changed as per requirement for each application. Figure 6.4 shows an example of part of the source code used for the mobile application shown in Figure 6.3. The Example shows the main page used for presentation of forecasting results for the Krishna-Bhima model (selected in the upper left corner on figure 6.3). The page uses a number of java utilities and legends collected in a folder included in javascripts.

Page 14

Figure 6.4 Example on some of the source code used for the example in Figure 6.3 for Mobile Devices.

Referring to the example, Figure 6.4, following features are included on the main page: Header Section Title on WEB page and inclusion of Google scripts. Data Section Inclusion of data and plots disseminated for the Krishna-Bhima model. All data are disseminated in one folder (PNG_Krishna1). For preparation of a new model WEB page these lines should be changed accordingly. Included Java scripts This section includes the Java utilities applied on the WEB page. (utilities, variables and legends) Initialize Google Maps Initialization of Google Maps Included KML layers Specification of KML-layers included on the map. Three KML files are included in the example: Outline of total catchment area, Outline of each catchment and the MIKE11 model setup. Included of KML files can be changed as per requirement.

Page 15

Include Objects This section includes specification for buttons, text and images include on the web page. Following objects are included in the example: - 2 buttons for selection of models (two river systems, if required). More buttons can be included as per requirement - 3 buttons for selection of data types (discharge, profiles and precipitation) - Button for zooming on the graph - Button for selection of data in a tabular view - Button for closing of a tabular view - 2 buttons for time step navigation - Text field definition for showing actual time step - Text field definition for Legend - Definition of table for tabular view of data - Definition for viewing a PDF file (not include on this web page) Finally all objects on the mobile WEB page are scaled and located according to the actual size of the WEB page. Scaling is performed via the java script utility: “mapstyle_mobile”. Setup of the web page The last section contains the setup of the actual web page in a tabular format: - First row : Header - Second row: Google Maps - Third row: selection of data types (3 types include in this example) - Fourth row: Graph text - Fifth row: Graph. A new model web pages can be prepared from this page shifting following sections: - Data Section: specification of model data - KML layers: specification of KML layers to be included - Objects: Setup of a model button for the actual model (in the example the first 2 buttons)

Page 16

7

RESERVOIR OPERATION MODULE Reservoir Operation can be performed via the Reservoir Operation Module. The tool is used to show the conditions for each reservoir included in the model setup and it can also be used for scenario simulations (as described in Chapter 8). The page consists of a Grahical View, as shown in figure 7-1, (water level left axis and inflow & outflow right axis), a Tabular View (time series of inflow, outflow and reservoir levels) and a Reservoir Overview on the bottom of the page. Figure 7-1 shows an example for the operation of the Warasgaon Reservoir. Future reservoir releases for the forecast period can be entered last column. After saving (right hand top button), the model is run again and a refreshed plot can be viewed. This is a trial and error simulation procedure for arriving a desired reservoir operation for the forecast period. The graph shows a 3 days period (1 day before Time of forecast, indicated with a grey vertical line and the forecasting period of 2 days). From Figure 7-1 it appears that the inflow (red line) is higher than outflow (green line) until ‘2006-07-30 03:00’, when the outflow exceeds the inflow. The water level (blue line) increases, therefore in the first part of the simulation, while it decrease when the ouflow eceeds the inflow.

7.1 Example of the Reservoir Operation Page.

Page 17

8

SIMULATION OF SCENARIOS Simulation of scenarios is activated when running RTSF & ROS in offline mode. The procedure to run a scenario for a specific forecast date is as follows: 1) If required change to offline mode, (see Figure 8.1, first red marking ) 2) Select the time of forecast for the scenario simulation in the Time of Forecast selection box (see Figure 8.1, second red marking).

Figure 8.1. RTSF & ROS Scenario Management: 1)Offline mode, 2) Selection of time of forecast, 3) Change of MIKE11 input data, 4) Scenarios group box.

Available dates for Time of Forecast, are the dates in the names of the model zip files saved in the archive folder (see Chapter 3, folder locations). As a part of the forecasting batch job (see Chapter 4), these zipped files are automatically prepared after finishing of the forecast simulation. It is also possible manually to copy or download a zipped forecast to the achieve folder. The latest forecast is always available as the first choice in the Time of Forecast selection box. After having selected the forecast date, the setup of a scenario simulation is prepared as follows: Reservoir Operation This features are described in Chapter 7.

Page 18

Rainfall Forecast Click on the Scenarios Rainfall Forecast-button (see Figure 8.1, red marking 4). Alternative Rainfall Forecasts are now specified via the Rainfall Forecast Editor as shown on figure 8.2.

Figure . 8.2. Rainfall Forecast Editor. Specification of new rainfall forecast.

It is possible to modify rainfall for the last 24 hours up to time of forecast and for the next 24, 48 and 72 hours for each sub-catchment as shown on the figure 8.2. If no new values are entered, the existing values are used in the simulation Other MIKE11 input data Alternative input data to MIKE11, can also be included in the scenario simulation. Click on the MIKE11 Editor-button (see Figure 8.1, red marking 3) to change MIKE11 input using the MIKE11 editor. As an example, it may be required not to apply data assimilation at all locations. To disconnect the data assimilation at a location: Uncheck the “Apply error forecast”checkbox at the location using the data assimilation editor as shown in Fig. 8.3.

Figure 8.3. Disconnection of data assimilation at a location for a scenario simulation.

Page 19

Start New Simulation After the alternative scenario has been specified, a new simulation can start when pressing the Start new simulation-button (see Figure 8.1, red marking 4). When the simulation is completed press the Refresh-button. Simulation results are now visible in the Graphical View (figure 8.1), where the scenario (black line) can be compared with the original simulation (blue line). In the Tabular View the 2 simulation results are shown in the 2 columns as, original: Sim and scenario: New (see figure 8.1). Saving and Plot Scenario After having completed a success-full scenario simulation, the setup for the actual scenario can be saved as a complete zipped model setup and saved in the archive folder. Plots of Results from the Scenario simulation will also be processed for presentation on a predefined WEB page. The saving of the Scenario is done through a predefined config file, which contains the number of the actual Scenario. Figure 8.4 shows an example, which has 5 predefined config files used for saving and plotting the scenario. The config file should contain the data location folder for presentation of the results (see Chapter 3).

Figure . 8.4. Selection of a config file for saving and plotting of a scenario.

After having saved the scenario it can then later be selected again from the Time of Forecast selection box (The prefix Scenario are added to the time of forecast to identify these simulations). Disseminate Scenario The last Scenario tool is Disseminate Scenario, used for the dissemination of a scenario. This tool will initiate a predefined batch job included in the config file for the actual scenario used to disseminate the scenario (eg. Upload of a scenario to a WEB page).

Page 20

9

DISSEMINATION OF FORECASTS

The tools for the dissemination of forecast results are being developed and therefore, this functionality is not implemented in the present version. The tools will include dissemiantion of various forecast results via E-mail and sms to pre-selected organisations and individuals.

Page 21