So long as you get the GIST of it
  1. Classes
  2. /
  3. Remote Sensing -- GEOG526
  4. /
  5. Labs
  6. /
  7. Lab - INTRODUCTION TO MODIS TIME SERIES

Lab - INTRODUCTION TO MODIS TIME SERIES

Learning Objective

This lab serves as an introduction to MODIS data products, which are widely used in a variety of terrestrial, aquatic and atmospheric remote sensing research and applications. This lab will familiarize you with MODIS vegetation indices products and various land cover/land use interpretation tasks. In addition, you will become familiar with the characteristics of MODIS imagery and time-series.

Outline:

Submission requirements

Materials (click to download)

Data Name Description
GEOG111_Lab2Questions.docx Handout to turn in

Guide

Tutorial

Part1 and Part2 are optional, so there are no questions about these two parts. Part1. Download the MODIS data

  1. Go to the website https://lpdaac.usgs.gov/data_access, then you can find a list of links to access the data. Click the link of Reverb (Launch Reverb). We will be able to search and download the data from the Reverb system. (Or you can directly go to http://reverb.echo.nasa.gov). 1 choose the bounding box 2 Draw a box to select the area that you are interested. (Not necessary to be the same of our example) 3 Input your Key word: MODIS NDVI in the Search Terms section 4 Specify the start and the end of time. 5 Check the dataset you would like to download. In this example, we select MODIS/Aqua Vegetation Indices 16-Day L3 Global 250 m SIN Grid V005

  2. After finishing the five steps shown above, then you can go to the end of this page and click . You will see a new window about the data quality summaries for the dataset we chose before, “Accept” it.

  3. Three data files were selected. “Browse” is to show the quick view of NDVI and EVI. Click this button, you will see more details. After checking the basic information of these files, click to add them into cart (the color change from blue to yellow after you added them). At last, you click .

  4. Now you can download the data by clicking at the end of page. It will appear an window to download the paths of data files. Keep the default choices and click “Save”, and then you can choose a directory to save the *.txt file on your own path.

  5. Open the file you saved from the last step. And you can copy the path to the browser to download the files. Be patient, and it will take several minutes to finish the download. e.g.

Part2. Import the MODIS VI product in ERDAS IMAGINE 2014. Did you remember the data import process (import SPOT6 *.jp2 to *.img) from last lab? We will do similar steps. Go to “Manage Data” and choose “Import Data”. Change the “format” to HDF in the Import window. Then you can import the data to img file format. This process will take very long due to the large date size and heavy computation. You can stop this step. We will provide the data for later use.

PS. NASA has its own special tool for the MODIS data to do reprojection and mosaic, called MRT (MODIS Reprojection Tool). It is very fast and convenient; you can download and find further details in https://lpdaac.usgs.gov/tools/modis_reprojection_tool. We won’t cover it in this lab.

Please, read the lab guide before you start the following sections. Part 3. Basic information for MODIS

  1. Fill in the table below for MODIS (combining the information from the guide and the link https://lpdaac.usgs.gov/dataset_discovery/modis/modis_overview . For the band range name, you can use what you learned from the previous labs (compared to TM/SPOT) or the definition in http://en.wikipedia.org/wiki/Electromagnetic_spectrum . Remember to write values with units. (6 pts)

Band number Wave range Resolution Band range (e.g. red, blue, green, near-infrared, or infrared) 1 ______________ ______________ ______________ 2 ______________ ______________ ______________ 3 ______________ ______________ ______________ 4 ______________ ______________ ______________ 7 ______________ ______________ ______________ 31 ______________ ______________ ______________

  1. Read carefully about the MODIS Naming Conventions below, and then answer the questions. MODIS filenames follow a naming convention which gives useful information regarding the specific product. For example, the filename MOD09A1.A2006001.h08v05.005.2006012234657.hdf indicates: MOD09A1 - Product Short Name .A2006001 - Julian Date of Acquisition (A-YYYYDDD) .h08v05 - Tile Identifier (horizontalXXverticalYY) .005 - Collection Version .2006012234567 - Julian Date of Production (YYYYDDDHHMMSS) .hdf - Data Format (HDF-EOS) The MODIS Long Name convention also provides useful information. For example, all products belonging to the MODIS/Terra Surface Reflectance 8-Day L3 Global 500m SIN Grid V005 collection have the following characteristics: MODIS/Terra - Instrument/Sensor Surface Reflectance - Geophysical Parameter 8-Day - Temporal Resolution L3 - Processing Level Global - Global or Swath 500m - Spatial Resolution SIN Grid - Gridded or Not V005 - Collection Version

A) The vegetation indices product file name is MOD13A2.A2013209.h10v05.005.2013226035948.hdf Please answer the following questions. (hint, for the long name, you may search the short name on the LPDAAC website, and the title of its home page is the long name, it will very similar to the example format) (4 points)

Product short name ___________________________________ Product long name ___________________________________ Geophysical Parameter ___________________________________ Sensor name ___________________________________ Julian Date and the regular date of Acquisition ________________and _______________ Collection Version ___________________________________ Spatial resolution ___________________________________ Temporal resolution ___________________________________

B) The MODIS/Aqua Land Surface Temperature and Emissivity (LST/E) products, MYD11C2.A2004150.004.2005226210814.hdf (4 points) In order to answer the questions below, you may need to search through Internet. (This is the link for MYD11C2 products: https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/myd11c2 )

Product short name ___________________________________ Product long name ___________________________________ Geophysical Parameter ___________________________________ Sensor name ___________________________________ Julian Date and the regular date of Acquisition ________________and _______________ Collection Version ___________________________________ Spatial resolution ___________________________________ Temporal resolution ___________________________________

  1. Estimate the NDVI from the figures below (the percentages indicates the reflectivity, the equation for the NDVI is on the guide). The left tree is healthy and the right one has withered. Then compared the values between these two trees, and why NDVI is helpful to show the healthiness of vegetation. (2 (calculations) +1(the comparision) points)

  2. Estimate the NDVI from the spectral profiles. The figure shows the profiles of green vegetation, dry vegetation, and soil. First, draw and label the band ranges which are required to calculate NDVI (Using MODIS band range) in this figure. Then you approximate the band-averaged reflectance to calculate NDVI for each feature in the graph below. (Show ALL your work and think about what information NDVI provides.) (9 points)

A) Which bands are used for calculating NDVI and their spectral range (Wavelength, with units) ( 2 points)

Band Name Spectral range 1 _________________________________ _________________________________ 2 _________________________________ _________________________________

B) Draw the spectral ranges of these two bands on the figure (1 points) C) Estimate the reflectivity for each feature below and calculate the NDVI (4.5 points) (show ALL your work!)

Surface features Reflectivity 1 Reflectivity 2 NDVI Green Vegetation Dry Vegetation Soil

D) Compare the NDVI you calculated above, which feature is with high NDVI? Why NDVI can distinguish vegetation and non-vegetation? (1.5 points)

  1. From the G:/Fall-2015/GEO526, copy the folder of Lab09 to your local drive. Lab09/MYD13Q1 contains the original MODIS vegetation indices file (.hdf) and the reprojected files (.tif) exported from hdf file. Open ERDAS and load the tif files, display the images and answer the following questions.

First, carefully read about the vegetation products available, including the layer information. https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mod13q1 The units, bit types, and scale factors are essential for you to get the correct values in MODIS products.

Science Data Sets (HDF Layers) (12) UNITS BIT TYPE FILL* VALID RANGE ** SCALE FACTOR*** 250m 16 days NDVI NDVI 16-bit signed integer -3000 -2000, 10000 0.0001 250m 16 days EVI EVI 16-bit signed integer -3000 -2000, 10000 0.0001 250m 16 days VI Quality detailed QA Bits 16-bit unsigned integer 65535 0, 65534 NA 250m 16 days red reflectance (Band 1) Reflectance 16-bit signed integer -1000 0, 10000 0.0001 250m 16 days NIR reflectance (Band 2) Reflectance 16-bit signed integer -1000 0, 10000 0.0001 250m 16 days blue reflectance (Band 3) Reflectance 16-bit signed integer -1000 0, 10000 0.0001 250m 16 days MIR reflectance (Band 7) Reflectance 16-bit signed integer -1000 0, 10000 0.0001 250m 16 days view zenith angle Degree 16-bit signed integer -10000 -9000, 9000 0.01 250m 16 days sun zenith angle Degree 16-bit signed integer -10000 -9000, 9000 0.01 250m 16 days relative azimuth angle Degree 16-bit signed integer -4000 -3600, 3600 0.1 250m 16 days composite day of the year Julian day of year 16-bit signed integer -1 1, 366 NA 250m 16 days pixel reliability summary QA Rank 8-bit signed integer -1 0, 3 NA

*The values in the Fill column means that if the data are unavailable for a pixel (NoData), use the listed fill value and assign that pixel to that value. ** Valid range shows the meaningful values in the product. If the pixel value ranges between these two values, it is valid and can be converted to the real scale by multiplying by the “scale factor”. *** SCALE FACTOR is used to calculate the real scale pixel value. Multiplying valid range by the scale factor gives you the real scale value.

The next exercise is to calculate the real scale value for NDVI and EVI.

A) What’s the real range for NDVI and EVI (note, it is not the valid range for the NDVI and EVI product list the table above)? There are four pixels representing the dense vegetation, deep water, bare soil, and NoData in the image, what’s the real scale NDVI if the pixel values are -3000, 5600, -1050, and 1? (Show all your steps, 4 points) And assign the surface feature to each NDVI. A: dense vegetation ; B: deep water; C: bare soil; D: NoData Pixel values (NDVI) Scale factor The real scale NDVI Surface Features (A, B, C, or D), choose from the list above -3000 5600 -1050 1

B) Use the inquire tool under Home tab/Information section (or select the tool from the right click menu on the image). Navigate the cross to the location in the inquire window below and find the pixel values of reflectance (band 1(Red), 2(NIR), 3(Blue), and 7(MIR)), NDVI and EVI, and calculate the real scale values for those points. (be sure you are clear about the fill value, valid range and multiply by scale factor for each layer) (12 points).

For example, change the first choice to “File”, then you can type in the file coordinates provided in the window below, then read the FILE PIXEL value in the window.

Location 1: File Location (X,Y) -> (5777,2424) Pixel value (or File Pixel) Scale factor Real scale value NDVI EVI Red NIR BLUE MIR

Location 2: File Location (X,Y) -> (5300,2928) Pixel value (or File Pixel) Scale factor Real scale value NDVI EVI Red NIR BLUE MIR

C) Using the two tables above, use the pixel values in NIR and RED to calculate NDVI, and compare the values with the NDVI provided in the products. (2 points)

Location 1: Location 2:

  1. Think carefully about the differences between TM and MODIS. Discuss the pros and cons of the two datasets and their potential applications. (List at least 3 pros and cons and 2 suitable applications for each sensor) (5 points)

Part 4: MODIS NDVI Temporal Profiles In a viewer open the MODIS time-series located at: G:/Fall-2015/Geo526/Lab09/Terra/terra_2012_subset.img. NDVI in this file is still in the valid range (-2000 – 10,000). The subset is of an agricultural area in western Kansas.

  1. Each layer is a 16-day composite of NDVI values. Assign the Red, Green and Blue color guns to layers 8, 13 and 18, respectively. Open the Spectral Profile tool under the Utilities subtab under Multispectral. Use the crosshairs to create multiple spectral profiles in one graph for one field for each color below. Examine the NDVI profiles over the year. Using the table below showing time periods and Julian days for 16-day composites, label the crop type you think each color is and the time of year the peak or maximum NDVI occurs. (6pts)

Colors in MODIS Time-Series Crop Type Date of Maximum NDVI

  1. Red
  2. Royal blue
  3. White
  4. Green