Why Upgrade?
You have been using Windographer version 1 and you wonder whether to upgrade to version 2? Here are some reasons.
Better Quality Control
The centerpiece of version 2 is a powerful new quality control system that lets you highlight, classify, and filter data by applying flags to data segments. You can define any number of your own flags, apply them to any number of data segments either manually or automatically, and then use those flags as criteria to include or exclude data from virtually any of Windographer's calculations, graphs, and export files.
Flags
A flag is simply a category, defined by a name and a color. Windographer provides multiple ways to apply flags to data segments. When you apply a flag to a data segment, you are categorizing that data segment. If you apply the 'Icing' flag to a data segment, for example, you are identifying that data segment as having been affected by icing. What you do about that categorization is a separate issue, one that you can decide later. You might, for example, choose to exclude from calculations any data flagged with the 'Icing' flag, but to include data flagged with the 'Tower Shading' flag. Or you might want to color code the points in a scatter plot according to flag. Or you might simply want to see a table showing the number of data points in each data column that were flagged with each flag.
The process of flagging a data segment does no harm to the data itself, and is entirely reversible. Flagged data points remain visible in time series graphs, as in the example below that shows where the user has applied the 'Icing' flag to a 60-hour segment of the 'Speed 45m' data column:
So flagging a data segment does not remove or harm that data segment, it simply marks it as belonging to a particular category. But this categorization of data enables many powerful capabilities, the most important of which is the ability to filter by flag.
Defining Flags
By default, Windographer includes a list of several flags, one for icing, one for tower shading, one for low quality data, one for invalid data, and so on. So you do not need to create or modify any flags unless you choose to do so. You can manage your list of flags in the Define Flags window, which appears below. You can even build a list of your favorite flags. Windographer will make these flags available in every document you open in Windographer.
Applying Flags
Three windows let you apply flags. The Flag Data Manually window lets you flag data segments by clicking and dragging on a time series graph:
The Flag Data With Rules window, which appears below, lets you create and execute flag rules. A flag rule is a set of instructions that tell Windographer to apply a certain flag to certain data columns under certain conditions. When you execute a flag rule, Windographer searches through the data set and applies the specified flag wherever it finds those conditions.
A flag rule might, for example, apply the 'Low Temperature' flag to all wind speed sensors whenever the temperature drops below -15°C. Or it could apply the 'Tower Shading' flag to a particular wind speed sensor whenever the wind direction falls within a particular range.
The flag rule in the screenshot below tells Windographer to look for events in which any wind direction standard deviation column drops below 1.5° for twelve time steps or more, and the temperature drops below 1°C, and to apply the ‘Icing’ flag to all wind speed and wind direction data columns as a result.
The flagging capability is particularly powerful when applied to the common problem of tower shading. That’s because Windographer can automatically generate the flag rules necessary to flag tower shading events. All you have to do is open the Flag Tower Shading window, which appears below, and verify that Windographer has correctly identified the shading patterns. At the click a button, Windographer flags all tower shading events in the data set, at all measurement heights.
You can mix manual and automatic flagging in any way you wish. You could, for example, start by executing your favorite set of flag rules that might catch 80% of the problems in the data set in a few seconds. Then later when you have sufficient time, you could systematically step through the data set, visually reviewing and manually adjusting the work done by the flag rules. Alternatively, you might choose to rely entirely on manual flagging, or you might refine your flag rules to the point that you rely on them exclusively.
Filtering Data by Flag
The main reason to flag data segments is to enable you to filter them out of calculations and export files. Windographer makes this easy by presenting a list of checkboxes that correspond to the flags you have applied in the data set. An example appears below:
To include data segments flagged with the 'Icing' flag, you simply check the checkbox labeled 'Icing'. Otherwise, you do not check that checkbox. This list of checkboxes appears in almost every window of Windographer, from the Wind Rose tab to the Reports tab to the Turbulence Analysis window to the Export Data window. In each case, the window will react immediately as you check or uncheck these checkboxes, updating the graph, table, or export file preview to reflect your new filter settings. So you can very quickly see the effect of including or excluding the different categories of data.
Better Import
Version 2 can open many kinds of data files that version 1 can't, including NDF files written by the SecondWind Nomad2 data logger and NSD files (site files) written by the Symphonie Data Retriever software by NRG Systems. It also does a better job of opening RWD, Triton, ASC, ZephIR, and Windcube data files. Most importantly for anyone who stores data in a database, Windographer version 2 can import from and export to any SQL database using stored procedures.
Better Append
The new Append Data Set window shows very clearly where the new data will go, and gives you much more control over the append process:
Better Vertical Extrapolation
We renamed the Virtual Anemometer window to the Vertical Extrapolation window and added several options that allow you to specify the wind shear parameter by month, hour of day, direction sector, or any combination of these variables. The new window appears below.
This new flexibility lets you apply wind shear patterns from another data set when extrapolating. For example, you could extrapolate from a met tower using wind shear patterns measured from a nearby SoDAR data set.
Better Analysis Capabilities
Version 2 incorporates many new and improved analysis modules.
New Tower Distortion Analysis Module
This new window plots the ratio by direction of the outputs of all paired wind speed sensors, so you can see the shading patterns at a glance and spot problems such as misaligned booms. It also calculates for each sensor pair the tower distortion factor, which incidates the severity of tower distortion, and the scatter factor, which indicates the degree of consistency between the two anemometers.
New Inflow Angle Analysis Module
Version 2 recognizes vertical wind speed data and uses it to calculate inflow angle, which it plots versus speed, direction, month, or time of day:
New Short Time Interval Analysis Module
This new window allows you to see a snapshot of the vertical wind shear profile in a single time step, or averaged over several time steps. This lets you see the full complexity of the wind shear patterns, and it can help detect problems such as backscattering in SoDAR data sets.
Better Filtering
Almost every window in Windographer now includes a standard set of filter controls that let you filter by flag, date, direction sector, or any other data column. This set of filter controls appears highlighted in the screenshot of the Histogram tab below. As you make changes to these filter settings, Windographer immediately updates the graph or calculation you are looking at to reflect the new settings.
Better Export
We have improved the Export Data window in several ways to help you move data from Windographer to the next step of your analysis. It now exports to Meteodyn WT, and the TAB file export feature can now remove seasonal bias so that Windographer calculates the mean of monthly means for each wind speed and direction bin. This avoids seasonal bias in data sets that cover a non-integer number of years, such as an 18-month data set that covers two summers but only one winter.
