Practical exercises: Rainfall data
Transcript of Practical exercises: Rainfall data
1
Practical exercises: Rainfall data
1 Understand the basic layout of the Data and Information application
In this exercise, you will learn about the basic functionalities within the data and information application and
will learn more about the functionalities to display and download the TRMM data. Please note that this exercise
is using Ghana as an example of a focus area, but could be done with any other basin available in the Flood and
Drought portal.
1.1. Get started with the Data and Information application
Open the Flood and Drought portal through:
www.flooddroughtmonitor.com and log in.
The Data and Information application is accessible from the Home page of
the Flood and Drought portal – click on the icon seen to the right to enter
the Data and Information application.
The first time the Data and Information portal is opened, you will be presented with the below view
without any available data. From here you can start adding data sets you would like to work with.
1.2 Explore datasets available on the Data and Information application
Click on the ‘Data’ tab on the upper left side of the map to open list of datasets currently available.
• To add or remove datasets from your list, click on the ‘Select’ Function. It opens a library of available datasets as shown below.
2
• Select datasets you would like to work with (use search box where needed). To start with,
select:
o Rainfall (TRMM)
o Monthly ensemble (TRMM)
o Monthly mean (TRMM)
• Click on the “Update” button to make the selected data available for viewing and download
from your main application view.
Background information is available for each data type by clicking on the data set name.
3
4
2 Understand the main functionalities when working with data: TRMM dataset
This part of the exercise will provide you with an overview of the main functionalities when working with
datasets on the Data and Information application.
2.1 Get information on the dataset status
• The “Last Update” flag in the top ribbon indicates when was the update of data in the selected area. A green mark indicates that update has been done lest then 3 days ago, else an orange or red mark will be shown.
• Click on “Status” within the Data menu to get an overview of the last time stamp for the different data types
• Click on “GIS layers” to download shape files for the basin, sub-basin and any station data.
2.2 View and download TRMM raster data
Select TRMM data
• Click on “Select” within the Data menu to open dialog listing all the available data types
• Select the TRMM rainfall data
• Click “Update”
Notice that the data type now appears in the main data dialog. This means that the data is now
available for viewing, analysis and download.
View the TRMM data
• Select ‘Rainfall (TRMM)’ data in the dataset
list
Note that the data is now shown on the map
displaying data for the last available time step (see
timesteps on the bottom of the window)
• Look at other time steps by clicking on the
time line in the bottom
• Look at other time steps by typing another
date in the date box
Note the legend to the right showing the data
values for the different colours.
5
Download raster data
To download a raster data set:
• Select the “Raster File” tool
• Specify a start and end data
• Press ‘Download’
Please note: the download file size might be
very large if a long time-period is selected.
All raster data are downloaded as netcdf
files.
• Download the TRMM data for the
last year
• [optional] View the data using a GIS
software (such as QGIS) or other
freeware tools (Panoply NASA tool).
3 Data tools
This part of the exercise will illustrate the different options for visualising and downloading time series data.
Introduction to tools
After selecting a dataset in Data menu, the
associated tools for the specific data type
will be enabled in the tools section below.
Please note that the list of tools depends on
the selected data type!
3.1 Time series
There are several options for plotting time
series.
The available tools are:
• Time series: Area weighted time
series for a specific area
• Time series (monthly): Monthly
accumulated time series for a
specific area
• Table: statistical data for the data
type
Note: the area weighted time series is
calculated by calculating the overlap
between the area (polygon) and the
satellite raster data using the formula:
6
𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝑇𝑆 = ∑ 𝑣𝑎𝑙𝑢𝑒∗𝑓𝑟𝑎𝑐𝑡𝑖𝑜𝑛
𝑆𝑢𝑚 𝑓𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑠 ,
where fraction is the fraction of a data cell
within the area.
Area when working with time series
The time series could be plotted on a number of different
spatial scales:
• All focus area: Weighted time values for the whole
basin
• User location: Weighted values for a user specified
location
• Subarea layer: Weighted values for a selected sub-
basin
• Point layer: Weighted values for a selected station
(point location)
Help on tools
To get more information about a tool, click
on it and an information box will be
displayed
Time series exercises
Explore the options for showing and downloading time series data through following exercises:
• Select the TRMM rainfall data
• Select the Time series tool for “all focus area”
• Try the different zoom options (ask the trainer for advice if needed)
• Download the data and open the csv file in Excel or a text editor
• Do the same steps but use the tool called “Time series (monthly)”
• Note that the monthly time series also contains a time series representing the long term average
• Download and open the csv file in Excel or a text editor (units of the download file correspond to
units you see in the legend when displaying values on the portal map).
Compare time series
Use the compare tool function to compare time
series across the area
• Select time series (monthly) tool
• Select an area (for example in the northern
part of the area)
• Click on “new chart”
• Select another area (for example in the
southern part of the area)
7
• Click on “Overlay chart”
• Analyse the compared time series
3.2 Table tool
The table tool is available only for some
datasets. The purpose of the tool is to
display different types of statistical
information. The type of information will
depend on the specific tool.
For TRMM rainfall monthly accumulated
values are displayed.
• Try the Table tool using the TRMM
data
• Export the table values to Excel and
make an overview table
• Identify the driest and wettest
month on both data set
• Discuss the difference between the
data
3.3 Envelope plot This exercise will demonstrate the advanced plotting options “Envelope” and “Column chart”.
Envelope plot
The envelope tool will be available for all data
types consisting of several members (or
ensembles) as this plot type is a way to simplify
the presentations of many members.
The envelope plot consists of 3 lines: 25 and 75
percentile and the median:
• Median: The median of all members at all
time steps
• 25 percentile: Shows the value for which
25% of the members are below
• 75 percentile: Shows the value for which
only 25% of the members are above
8
The three lines are often used to represent the
most likely outcome of a data type consisting of
many members.
Apply envelope plot tool for the TRMM dataset
• Select the “Historical ensemble” (if not in the list of available datasets, then first click on
“Select” and add the TRMM historic ensemble data from the list – remember to click
‘Update’)
• Start by plotting a Time series (Monthly) from the Tools section – notice that the data type
consists of a number of members (click on ‘New Chart’ to view the plot)
• Plot the Envelope plot
• Discuss if you understand the envelope plot and if you are able to see the connection with
the monthly data. Ask the trainer if there are any questions.
• Discuss how the data could be used to evaluate the climate variability. Are you able to
identify periods or seasons with high or low variability.
3.4 Monthly mean (TRMM)
The monthly mean layer shows rainfall value
derived from daily rainfall dataset. The value is
given on the 15th of the month for the entire
month. Values are given in mm/day.
• Select monthly mean (TRMM) from the list
of data available.
• View monthly mean values for various time
steps and discuss the differences
• Download monthly mean data as a raster
file, and, if possible, view using QGIS or
other application.
9
Exercise: Rainfall Indices Rainfall indices overview
In this exercise, you will learn you will learn how the Data and Information application can be used to get an overview of the climate in a specific area or basin, with focus on rainfall data. Objective: learn how satellite-based data can provide valuable information regarding the climatology in an area, using rainfall data.
Open the Data and Information application
Open the Flood and Drought portal through
www.flooddroughtmonitor.com and log in.
The Data and Information application is accessible from the Home page
of the Flood and Drought portal – click on the icon seen to the right to
enter the Data and Information application.
Please refer to the introduction exercise to the Data and Information
application for a detailed description of the functionality.
1 Rainfall indices: Effective Drought Index (EDI) (TRMM)
The objective with this exercise is to demonstrate how different rainfall data sources could be used to provide an overview of the current and historic climate in the area.
1.1 EDI: Time series tool
The Effective Drought Index (EDI) is a measure for how the
weighted rainfall during the past 365 days compares to the long
term normalised mean.
If the EDI is below -1 then there would normally be a drought.
Consecutive days with drought would increase the stress on the
crops.
• Select EDI under TRMM data products
• View data as Time series
• Zoom in to the years with extreme dry conditions by
selecting the area on the chart.
• Click on the options in the chart (1m, 6m, 1y) to see the data
for different time steps.
10
• Change Area selection to show data on Subarea layer.
Explore the time series data for various subareas of the main
dataset.
1.2 EDI: Column chart tool
The column chart is a plot type only available for indices as it
represents the area covered by a specific index category at a
specific time.
• Select EDI under TRMM data products.
• View data as Column chart.
• Zoom in to the years with extreme dry conditions by
selecting the area on the chart.
• See column charts for various subarea layers and discuss
differences.
• Discuss how the rainfall distribution varies within the year.
Are there a distinct dry and wet seasons and how is the
variability of the seasons from year to year?
Note: The column chart shows how large part of the area is
represented by a specific EDI category. This plot could be used
to identify periods with a large coverage of very dry or very wet
rainfall conditions. Ask the trainer if you need more knowledge
of this plot type.
1.3 EDI: Combine time series charts
11
The time series tools allow for overlay of several charts to enable comparison between conditions in
different subareas layers.
• View EDI data using Time series tool and select data display under ‘Area’ to be by subarea layer.
• Select one subarea layer and create ‘New chart’
• Select another subarea from the list.
• Click on ‘Overlay chart’ function.
• The chart now shows the overlay of time series for various regions (see the dry and wet data on
the map and compare time series for respective regions.
• Discuss what does the data show for the respective years and how the two areas compare.
Discussion in groups
Discuss in groups how you would like the climate data for a specific are to be presented?
Should it be an overview table with the key climate data or should it be a report with tables and
figures?
2 Rainfall drought index compared to data
The objective with this exercise is to highlight the benefit of using drought index instead of data.
2.1 Using rainfall time series only
Identify a period of drought using the Rainfall (TRMM) resampled to monthly data
• View the monthly times series of Rainfall (TRMM) for a selected area
• Identify a period when a drought spell occurred
12
2.2 Using climatology data
Identify a drought event using the historical ensemble (climatology data)
• View the historical ensemble for the same area as above
• Identify a period when a drought spell occurred
2.3 Using drought index
Identify a drought event using the EDI dataset
• View the historical ensemble for the same area as above
• Identify a period when a drought spell occurred
2.4 Discuss the datasets and tools
• Discuss which dataset is more tailored to identify drought spell?
• Discuss pro and cons of the different methods
13
Exercise: Soil Moisture Index
Soil Moisture Index
This part of the exercise will introduce the soil moisture index as a data source for evaluation of the water
availability in the root zone. Please note that this exercise was developed using Ghana as an example but could
be done with any other basin available in the Flood and Drought.
1 Open the Data and Information application
Open the Flood and Drought portal through
www.flooddroughtmonitor.com and log in.
The Data and Information application is accessible from
the Home page of the Flood and Drought portal – click on
the icon seen to the right to enter the Data and
Information application.
Please refer to the introduction exercise to the Data and
Information application for a detailed description of the
functionality.
2 Soil Water Index (SWI)
This part of the exercise will introduce the soil moisture index as a data source for evaluation of the water availability in the root zone.
SWI is used as an indicator for the water availability in the upper part of the root zone.
SWI in itself is NOT a drought index but could be used to detect changes in the soil moisture pattern from year to year.
The soil water index (SWI) product contains daily synthesis of Soil Water Index derived from ASCAT SSM data at 25 km resolution (then resampled to 0.1 degree). The SWI algorithm uses an infiltration model describing the relation between surface soil moisture and profile soil moisture as a function of time. The algorithm is based on a two-layer water balance model. The data value used are using the largest characteristic time length giving the deepest penetration within the soil layers, and the values represent an average soil water index across 0.5 to 1 meter of the topsoil.
2.1 Make SWI data available
• Open selection window and add ‘SWI’ and ‘SWI
Percentile’ to the data layers list
• Read the data descriptions to understand how the
data is developed
14
2.2 Evaluate the raw SWI value
• Select SWI in the data layer list.
• Try selecting different time steps on the bottom of
the page to see the change in SWI over a season.
• Select Time series tool a and create a chart for the
focus area
• Try creating charts and overlaying charts for various
subarea layers to compare SWI values for same
season
2.3 Comparison with long term mean
To evaluate the impacts of soil moisture status and changes, you need to look at how the soil
moisture is expected to be under normal conditions. Deviations from long term mean may signal
changes in conditions, e.g. risk of drought.
This could be done by comparing the area data with the long-term mean data using SWI Time series
chart tool.
• Compare the values for the current year (1y) with the long-term mean (long-term mean is
displayed in red in the time series plot).
• Evaluate how the current season compares to the long-term average.
• Look at All years on record in the chart, compared to long term mean. Identify years where
the soil moisture seems to be significantly lower or higher than the average.
• Download data as an excel file. Identify for which years the SWI has been significantly below
long term mean.
15
3 SWI Percentile
The SWI percentile is an index illustrating how often the current soil moisture condition appears within the context of historic conditions for the same period. This is used to assess whether the soil moisture conditions are close to what would be expected (occurring frequently) or if they are rare (not occurring frequently).
• A percentile value of 95% means that 95% of the recorded values are below the current one (this can be interpreted as a very wet period)
• A percentile value of 5% means that only 5% of the historic values are below the current one (this can be interpreted as a very dry period)
• A drought or water scarcity is often defined when the soil moisture percentile drops below 30 or 20 %.
3.1 SWI Percentile assessment for selected area
We will use the SWI percentile to identify
areas and periods of time affected by
dryer conditions than to be expected.
• Select SWI percentile from the data
layers list
• Browse through different periods in
time using bar for time steps on the
page
• Discuss: are there areas where the
moisture conditions are much below
the historic values?
• Discuss: Are there areas where it is
very wet compared to historic
conditions?
• Compare the two maps for same
period in time, using SWI data vs SWI
percentile. What are the differences?
16
3.2 SWI percentile chart tool
The column chart tool expresses the coverage of a specific
index category within an area for the selected time
period. It can be used to identify periods with a high
coverage of severe index categories within an area.
• Select SWI percentile and then Column chart tool
• Create chart and discuss which years and time
period show signs of potential drought conditions
• Download data as an Excel file
• Identify historic events with extreme wet or dry
conditions using the excel file sheet.
• Can you see the results from Excel sheet in the
column chart?
17
Exercise: Vegetation Index
Vegetation index
In this exercise, we introduce the workflow in a drought assessment process with focus on identification and
location of drought hazards using a vegetation index. Vegetation and especially agricultural crops are often one
of the key targets in a drought assessment, as the food production will be one of the areas impacted by drought.
This part of the exercise will introduce different types of indices used to evaluate the impact on crops and
vegetation in general.
Please note that this exercise was developed using Ghana as an example but could be done with any other basin
available in the Flood and Drought.
Open the Data and Information application
Open the Flood and Drought portal through
www.flooddroughtmonitor.com and log in.
The Data and Information application is accessible from the Home
page of the Flood and Drought portal – click on the icon seen to the
right to enter the Data and Information application.
Please refer to the introduction exercise to the Data and Information
application for a detailed description of the functionality.
1 NDVI (Normalised Difference Vegetation Index)
NDVI is used for evaluating the drought impact on vegetation, where the change in NDVI values is
used as an indicator for vegetation impact. When comparing the current NDVI with long term
statistical values it is possible to evaluate the impact on the vegetation.
Resolution: Spatial: 5.6 km to 250 m, Temporal: 16-day composite
Satellite: MODIS (Terra), Data from 2000 – present, Global coverage
Description: Correlation with leaf area index and biomass, cloud sensitive
1.1 Add NDVI data
We will use the NDVI data for the exercise so
make sure that this data type is made available. If
not on the list, open and add the NDVI (5600m)
dataset to the available layers list.
1.2 Evaluate the raw NDVI data values
18
The NDVI values are closely related to biomass
and leaf area. NDVI is an indicator for the
vegetation mass in an area.
• Select NDVI (5600m) layer in the data list
• Select different time steps to see the
change in NDVI over a season.
1.3 NDVI time series tool
Select the Time series tool for NDVI data.
• Use the time series tool and time step
selection to evaluate the NDVI values over
time.
• Evaluate if you see a different seasonal
pattern in different parts of the basin.
• Select time series for subarea layer
• Compare NDVI values for different parts of
the basin by overlaying charts
• Discuss if the time series reflect what you
see on the map for same time periods.
Comparison with long term mean
To evaluate the impact on the vegetation you
need to look at how the vegetation is expected
to be. This could be done by comparing with the
focus area data with the long-term mean.
• Compare the values for the current year
with the long-term mean (long-term mean
is displayed as a line red in the time series
plot).
• Evaluate how the current season compares
to the long-term average. Identify years
where the NDVI value seems to be
significant lower than the average (you will
need this analysis for the next exercise).
19
2 Vegetation Condition Index (VCI)
The Vegetation Condition Index (VCI) compares the current NDVI to the range of values observed in
the same period in previous years. The VCI is a commonly used drought index and is expressed in %.
It evaluates how the current vegetation growth compares to the minimum and maximum values
observed for this specific area in the previous years.
𝑉𝐶𝐼 =𝑁𝐷𝑉𝐼 − 𝑁𝐷𝑉𝐼𝑚𝑖𝑛
𝑁𝐷𝑉𝐼𝑚𝑎𝑥 − 𝑁𝐷𝑉𝐼𝑚𝑖𝑛
High VCI -> NDVI close to historic maximum
Low VCI -> NDVI close to historic minimum
2.1 Add VCI data
Make the VCI product available from data selection
list.
2.2 Evaluate VCI status
Evaluate the current status of VCI is in the selected
focus area.
• Are there areas where the vegetation growth is
close to historic maximum? (corresponding to VCI
value close to 100%)
• Are there areas where the vegetation growth is
close to the historic minimum? (corresponding to
VCI value close to 0%)
• Download the data as an excel file.
• Looking back at historic record, are you able to
identify periods with a significant reduced crop
growth?
• How does this compare with the periods identified
in the previous exercise?
2.3 Evaluate the coverage of a hazard
The column chart tool expresses the coverage of a
specific index category within an area, and is used to
identify times with a high coverage of severe index
categories within an area.
Select column chart tool from the tools list to view the
results.
Note: Some periods will show a coverage less than
100% - often during the rainy season. This is due to
cloud sensitivity of NDVI.
20
• Identify historic events with high and low
vegetation condition.
• How does the column chart compare to the
above analysis of the NDVI values?
• Compare the VCI data with rainfall related
data for same periods – are there any
correlations?