Collection 6.

1 MODIS Burned Area Product User’s Guide

Version 1.1

Louis Giglio
Michael Humber
Joanne V. Hall
Fernanda Argueta
University of Maryland
Luigi Boschetti
University of Idaho
David Roy
Michigan State University

November 2022
 Technical Contacts

Topic Contact
Algorithm and HDF product Louis Giglio ([email protected])
GeoTIFF and Shapefile product Michael Humber ([email protected])
Product validation Luigi Boschetti ([email protected])

Abbreviations and Acronyms

BA Burned Area
CMG Climate Modeling Grid
EOS Earth Observing System
EOSDIS EOS Data Information System
GeoTIFF Georeferenced Tagged Image File Format
HDF Hierarchical Data Format
LP-DAAC Land Processes Distributed Active Archive Center
MODIS Moderate Resolution Imaging Spectroradiometer
SDS Science Data Set
QA Quality Assessment

Document Change History

Version Date Description

1.0 5/2021 Original version adapted from C6 guide, version 1.3.
1.1 11/2022 Updated to reflect completion of C6.1 reprocessing and for ENVI 5.5.

Title page image: MCD64A1 cumulative area burned in the Central African Republic and South Sudan
during the 2004–2005 burning season.

2
Contents
1 Introduction 6
1.1 Summary of Collection 6.1 Algorithm and Product Changes . . . . . . . . . . . . . . . . . 6
1.1.1 Reprocessing Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.1 Granules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.2 Tiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.3 Collections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2.4 Climate Modeling Grid (CMG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 MCD64A1 Algorithm Summary 9

3 MCD64 Product Suite 9

3.1 Level 3 Monthly Tiled Product: MCD64A1 . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.1 Naming Convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.2 Data Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.3 Metadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1.4 Example Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 GeoTIFF subset for GIS visualization and analysis: MCD64monthly . . . . . . . . . . . . . 17
3.2.1 Naming Convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.2 Example Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 Shapefile subset for GIS visualization and analysis: MCD64monthly . . . . . . . . . . . . . 19
3.3.1 Naming Convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4 MCD64CMQ Climate Modeling Grid Burned Area Product . . . . . . . . . . . . . . . . . . 20
3.4.1 Naming Convention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.2 Global Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.3 Data Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.4 Metadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4.5 Example Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

4 Obtaining the MODIS Burned Area Products 24

4.1 Downloading the products from the fuoco SFTP server . . . . . . . . . . . . . . . . . . . . 24
4.1.1 HDF Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1.2 GeoTIFF files and Shapefiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1.3 Example sftp command line session . . . . . . . . . . . . . . . . . . . . . . . . . 26

5 Working with the product in ENVI 5.5 27

5.1 MCD64A1 (HDF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.2 MCD64monthly (GeoTIFF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.3 MCD64monthly (Shapefile) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

6 Working with the product in ArcGIS 30

6.1 MCD64A1 (HDF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.1.1 ArcGIS Pro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.1.2 ArcGIS ArcMap Desktop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3
 6.2 MCD64monthly (GeoTIFF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.2.1 Area of Interest (AoI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.3 MCD64monthly (Shapefile) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

7 Validation of the MODIS Burned Area Product 35

8 Known Problems 35
8.1 Cropland Burning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8.2 Pre-November 2000 Mission Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8.3 June 2001 Terra MODIS Outage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8.4 August 2020 Aqua Outage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

9 Frequently Asked Questions 36

10 References 37

11 Relevant Web and FTP Sites 38

Appendix A Coverage of the GeoTIFF subsets 39

Appendix B Coordinate conversion for the MODIS sinusoidal projection 40

B.1 Forward Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
B.2 Inverse Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
B.3 Applicability to 250-m and 1-km MODIS Products . . . . . . . . . . . . . . . . . . . . . . 41

List of Tables
1 Sizes of grid cells in Level 3 tiled MODIS sinusoidal grid. . . . . . . . . . . . . . . . . . . 8
2 Day-of-year of the first day of each calendar month. . . . . . . . . . . . . . . . . . . . . . . 10
3 MCD64A1 metadata stored as standard global HDF attributes. . . . . . . . . . . . . . . . . 12
4 MCD64CMQ metadata stored as standard global HDF attributes. . . . . . . . . . . . . . . . 21
5 Regions and bounding coordinates of the GeoTIFF subsets. . . . . . . . . . . . . . . . . . . 39

List of Figures
1 Collection 6 and Collection 6.1 MCD64A1 monthly global burned area time series. . . . . . 7
2 Scatter plots of per-tile Collection-6.1 versus Collection-6 burned area. . . . . . . . . . . . . 7
3 MODIS tiling scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 Coverage of the GeoTIFF subsets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5 ENVI 5.5 example Dataset Browser. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6 ENVI 5.5 Classic example Dataset Browser and image display. . . . . . . . . . . . . . . . . 28
7 ENVI 5.5 Shapefile import example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
8 Importing HDF data into ArcGIS Pro. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
9 Selecting the HDF Burn Date layer in ArcGIS Pro. . . . . . . . . . . . . . . . . . . . . . . 31
10 January 2021 MCD64A1 Burn Date layer for MODIS tile h16v07 displayed in ArcGIS Pro. 32
11 ArcGIS Pro export raster window. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

4
12 ArcGIS ArcMap export raster window. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
13 MCD64monthly Shapefile superimposed over Sentinel-2 image. . . . . . . . . . . . . . . . 34
14 Representative Direct Broadcast coverage during 2020 Aqua outage. . . . . . . . . . . . . . 36

5
1 Introduction
This document contains the most current information about the Collection 6.1 Moderate Resolution Imaging
Spectroradiometer (MODIS) Burned Area product suite. It is intended to provide the end user with practical
information regarding the use (and misuse) of the products, and to explain some of the more obscure and
potentially confusing aspects of the burned area products and MODIS products in general.

1.1 Summary of Collection 6.1 Algorithm and Product Changes

Initially driven by calibration issues that impacted the MODIS atmosphere products, a Collection 6.1 “mini-
reprocessing” was initiated in 2018 and later expanded to include the MODIS land products. For the land-
product suite the changes under Collection 6.1 are largely confined to the MODIS Level-1B calibrated
radiance products (e.g., MOD021KM) from which the various land products are derived.
The MODIS MCD64A1 burned area product is among those for which neither algorithm nor prod-
uct changes were made for Collection 6.1, and as such the Collection 6.1 MCD64A1 product is largely
unchanged from its Collection-6 predecessor (Figure 1). The small differences that occur across the two
collections are the result of the following changes:
1. The product is generated using the Collection 6.1 MODIS surface reflectance and active fire products
as inputs. The former in particular changed slightly as a result of 1) the changes made to the upstream
Collection-6.1 MODIS Level-1B calibrated radiance products, and 2) minor bug fixes made to the
surface reflectance production software.
2. For Collection 6, during some product months not all of the required MODIS active-fire input files
were properly staged within MODAPS for use as inputs to the MCD64A1 production software. For
Collection 6.1 this minor error was corrected, leading in some instances to a few extra days of active
fire observations at the beginning of the daily time series considered by the MCD64A1 mapping
algorithm.
3. The initially delivered Collection 6 MCD64A1 production software contained a bug that caused
burned grid cells in columns 0 and/or 2399 (the left-most and right-most columns) of 26 specific
MODIS land tiles to be incorrectly relabeled as unburned. A bug fix was implemented in September
2018 but applied only to the forward processing stream. With Collection 6.1 this problem has been
resolved for the entire mission period.
The net result of the above changes is that the Collection 6 and Collection 6.1 MCD64A1 products are
sufficiently consistent to be intermixed for many purposes. This consistency can be seen in the right panel of
Figure 2, where we have selected the MCD64A1 product for August 2003 as a representative example. The
sole exception to the otherwise excellent agreement occurred for the November 2000 product, which proved
to be an isolated anomaly. We traced the atypical discrepancies observed for this month to the Collection 6.1
MODIS surface reflectance product, which is an input to the MCD64A1 mapping algorithm. It is not yet
clear why the Collection-6.1 surface reflectance product differed more significantly for just this particular
time period.

1.1.1 Reprocessing Schedule

The Collection-6 MCD64A1 product will remain in forward production through December 2022 while the
Collection-6.1 reprocessing of historical MODIS data is underway. After this date forward production of all
MODIS Land products (including MCD64A1) will continue via the Collection 6.1 reprocessing exclusively.

6
 80 C6.1
Area Burned (Mha/month)

C6
60
40
20
0

2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021
Year

Figure 1: Collection 6 (blue dotted lines) and Collection 6.1 (red solid lines) MCD64A1 monthly global
burned area time series for the first 21-years of the MODIS data record.

November 2000 August 2003

16

16
12

12
C6.1 Burned Area (Mha)

C6.1 Burned Area (Mha)

6 8

6 8
4

4
2

2
1

1
0

0 1 2 4 6 8 12 16 0 1 2 4 6 8 12 16
C6 Burned Area (Mha) C6 Burned Area (Mha)

Figure 2: Scatter plots of per-tile Collection-6.1 (C6.1) versus Collection-6 (C6) MCD64A1 total burned
area for November 2000 (left panel) and August 2003 (right panel). Each circle represents a tile (e.g.,
h08v05) on the MODIS sinusoidal grid during the specified calendar month. The gray and light blue lines
indicate the 1:1-reference and least-squares regression lines, respectively. The August 2003 example is
highly representative of the Collection-6.1 product in general. In contrast the November 2000 case is an
isolated anomaly for which the differences are much more pronounced.

7
1.2 Terminology
1.2.1 Granules
A granule is an unprojected segment of the MODIS orbital swath containing about 5 minutes of data.
MODIS Level 0, Level 1, and Level 2 products are granule-based.

1.2.2 Tiles
MODIS Level 2G, Level 3, and Level 4 products are defined on a global 250-m, 500-m, or 1-km sinusoidal
grid (the particular spatial resolution is product-dependent). Because these grids are unmanageably large in
their entirety (43200 × 21600 pixels at 1 km, and 172800 × 86400 pixels at 250 m), they are divided into
fixed tiles approximately 10◦ × 10◦ in size. Each tile is assigned a horizontal (H) and vertical (V) coordinate,
ranging from 0 to 35 and 0 to 17, respectively (Figure 3). The tile in the upper left (i.e. northernmost and
westernmost) corner is numbered (0,0).

Figure 3: MODIS tiling scheme.

Note that the Level 3 MODIS products generated on the MODIS sinusoidal grid are colloquially referred
to as having “1 km”, “500 m”, and “250 m” grid cells. The exact cell sizes are shown in Table 1.

Table 1: Sizes of grid cells in Level 3 tiled MODIS sinusoidal grid.

Colloquial Size Actual Size (m)

“1 km” 926.62543305
“500 m” 463.31271653
“250 m” 231.65635826

8
1.2.3 Collections
Reprocessing of the entire MODIS data archive is periodically performed to incorporate better calibration,
algorithm refinements, and improved upstream data into all MODIS products. The updated MODIS data
archive resulting from each reprocessing is referred to as a collection. Later collections supersede all earlier
collections.
Neither the MCD45A1 nor MCD64A1 MODIS burned area product was produced in Collections 1
through 4. Both products were produced for the first time as part of Collection 5. For Collection 6 the
MCD45A1 burned area product was retired and replaced with MCD64A1.

1.2.4 Climate Modeling Grid (CMG)

MODIS Level 3 and Level 4 products can also be defined on a coarser-resolution climate modelling grid
(CMG). The objective is to provide the MODIS land products at consistent low resolution spatial and tem-
poral scales suitable for global modeling. In practice, there is a fair amount of variation in the spatial and
temporal gridding conventions used among the MODIS land CMG products.

2 MCD64A1 Algorithm Summary

The MCD64 burned-area mapping approach employs 500-m MODIS imagery coupled with 1-km MODIS
active fire observations. The hybrid algorithm applies dynamic thresholds to composite imagery generated
from a burn-sensitive vegetation index (VI) derived from MODIS short-wave infrared channels 5 and 7, and
a measure of temporal texture. The VI is defined as
ρ5 − ρ7
VI = ,
ρ5 + ρ7
where ρ5 and ρ7 are respectively the band 5 and band 7 atmospherically corrected surface reflectance.
Cumulative active fire maps are used to guide the selection of burned and unburned training samples and to
guide the specification of prior probabilities. The combined use of active-fire and reflectance data enables
the algorithm to adapt regionally over a wide range of pre- and post-burn conditions and across multiple
ecosystems. See Giglio et al. (2018) for a complete description of the algorithm.
The mapping algorithm ultimately identifies the date of burn, to the nearest day, for 500-m grid cells
within the individual MODIS tile being processed. The date is encoded in a single data layer of the output
product as the ordinal day of the calendar year on which the grid cell burned (range 1-366), with a value
of 0 for unburned land grid cells and additional special values reserved for missing-data and water grid
cells. The output product contains additional data layers for diagnostic purposes and to facilitate uncertainty
propagation into downstream products derived from the burned area maps, such as emissions estimates.

3 MCD64 Product Suite

Three different versions of the MODIS burned area product are available:
• The official MCD64A1 product in HDF-EOS format, which is available as part of the MODIS suite
of global land products.
• The re-projected monthly GeoTIFF version available from the University of Maryland.
• The re-projected monthly Shapefile version available from the University of Maryland.

9
3.1 Level 3 Monthly Tiled Product: MCD64A1
The MCD64A1 Burned Area Product is a monthly, Level-3 gridded 500-m product containing per-pixel
burning and quality information, and tile-level metadata.

3.1.1 Naming Convention

The file naming convention, which has been adopted by all standard MODIS products, is as follows:

MCD64A1.AYYYYDDD.hHHvVV.061.PPPPPPPPPPPPP.hdf

where
YYYY = year mapped
DDD = start day-of-year (ordinal day) of calendar month in which burns have been mapped (Table 2)
HH = horizontal tile coordinate on MODIS sinusoidal grid
VV = vertical tile coordinate on MODIS sinusoidal grid
061 = collection identifier
PPPPPPPPPPPPP = production date1

Example: The product file MCD64A1.A2006244.h31v10.061.2021106034604.hdf contains the

September 2006 burned area map for MODIS tile h31v10, located in northern Australia.

Table 2: Day-of-year (DOY) of the first day of each calendar month. The DDD field (see above) in the file
names of the MCD64 products will always have one of the 22 unique values shown here.

Non-Leap Year Leap Year

Month Start DOY Start DOY
January 1 1
February 32 32
March 60 61
April 91 92
May 121 122
June 152 153
July 182 183
August 213 214
September 244 245
October 274 275
November 305 306
December 335 336

1
This naming convention ensures that files always have a unique name: if a tile is reprocessed the last number, indicating the
day and time in which the file was processed, will be different, thus avoiding any confusion with obsolete data.

10
3.1.2 Data Layers
The product contains five data layers (Burn Date, Burn Date Uncertainty, QA, First Day, and Last Day),
each stored as a separate HDF4 Scientific Data Set (SDS).

Burn Date: Ordinal day of burn (1-366) for each 500-m grid cell, with 0 = unburned land, -1 = unmapped
due to insufficient data, and -2 = water.

Burn Date Uncertainty: Estimated uncertainty in date of burn, in days. Unburned and unmapped grid
cells will always have a value of 0 in this layer.

QA: 8-bit quality assurance bit field.

bit 0: 0 = water grid cell, 1 = land grid cell.

bit 1: Valid data flag (0 = false, 1 = true). A value of 1 indicates that there was sufficient valid data
in the reflectance time series for the grid cell to be processed. (NB. Water grid cells will always
have this bit clear.)
bit 2: Shortened mapping period (0 = false, 1 = true). This flag indicates that the period of reliable
mapping does not encompass the full one-month product period, i.e., burns could not be reliably
mapped over the full calendar month.
bit 3: Grid cell was relabeled during the contextual relabeling phase of the algorithm (0 = false, 1 =
true).
bit 4: Spare bit set to 0.
bits 5–7: Special condition code reserved for unburned grid cells. This code provides an explanation
for any grid cells that were summarily classified as unburned by the detection algorithm due to
special circumstances.
0: None or not applicable (i.e., burned, unmapped, or water grid cell).
1: Valid observations spaced too sparsely in time.
2: Too few training observations or insufficient spectral separability between burned and un-
burned classes.
3: Apparent burn date at limits of time series.
4: Apparent water contamination.
5: Persistent hot spot.
6: Reserved for future use.
7: Reserved for future use.

First Day and Last Day: The First Day and Last Day SDSs indicate the first and last days, respectively,
on which changes could be reliably detected within the time series, on a per-grid-cell basis. Note
that during periods of persistent cloud cover or frequent data loss, these dates will often lie within the
calendar month that was mapped, indicating that reliable mapping could not be achieved over the full
month.

11
3.1.3 Metadata
As with standard MODIS products, the MCD64A1 product contains global metadata stored in the ECS
CoreMetadata.0 and ArchiveMetadata.0 global attributes. For convenience, a subset of these metadata as
well as additional metadata are stored as standard HDF global attributes (Table 3).

Table 3: MCD64A1 metadata stored as standard global HDF attributes.

Attribute Name Description

BurnedCells Number of burned 500-m land grid cells during product month.
MissingCells Number of 500-m land grid cells during product month having too
few valid observations for algorithm to function.
LandCells Number of 500-m land grid cells in tile.
ValidLandCells Number of 500-m land grid cells for which mapping could be per-
formed.
ProductStartDay Ordinal start day of product calendar month (range 1 – 336).
ProductEndDay Ordinal end day of product calendar month (range 31 – 366).
year Product year.
tile Six-character tile descriptor, e.g., “h20v11”.

12
3.1.4 Example Code
Example 1: IDL code to read a global attribute and two SDSs in the MCD64A1 product.

mcd64a1_file = ’MCD64A1.A2006213.h20v11.061.2021308225836.hdf’

sd_id = HDF_SD_START(mcd64a1_file, /READ)

; read "BurnedCells" attribute

attr_index = HDF_SD_ATTRFIND(sd_id, ’BurnedCells’)
HDF_SD_ATTRINFO, sd_id, attr_index, DATA=burned_cells

; convert 1-element vector to scalar

burned_cells = burned_cells[0]

index = HDF_SD_NAMETOINDEX(sd_id, ’Burn Date’)

sds_id = HDF_SD_SELECT(sd_id, index)
HDF_SD_GETDATA, sds_id, burn_date
HDF_SD_ENDACCESS, sds_id

index = HDF_SD_NAMETOINDEX(sd_id, ’QA’)

sds_id = HDF_SD_SELECT(sd_id, index)
HDF_SD_GETDATA, sds_id, QA
HDF_SD_ENDACCESS, sds_id

HDF_SD_END, sd_id

help, burned_cells, burn_date, QA

The code produces the following output:

BURNED_CELLS LONG = 101302
BURN_DATE INT = Array[2400, 2400]
QA BYTE = Array[2400, 2400]

13
Example 2: Python code to read the Burn Date and QA SDSs in the MCD64A1 product.

import gdal
import numpy as np

in_file = "MCD64A1.A2006244.h31v10.061.2021106034604.hdf"

hdf_handle = gdal.Open(in_file)

# Print a list of all of the datasets within the hdf file

sds_list = hdf_handle.GetSubDatasets()
for sds in sds_list:
print(sds)

# Open a subdataset "N" as an array with sds_list[N][0], values of N are:

# 0 = ’Burn Date’, 1 = ’Burn Date Uncertainty’, 2 = ’QA’,
# 3 = ’First Day’, 4 = ’Last Day’
burndate_handle = gdal.Open(sds_list[0][0])
burndate_data = burndate_handle.ReadAsArray()

qa_handle = gdal.Open(sds_list[2][0])
qa_data = qa_handle.ReadAsArray()

# Array can be handled with Numpy now

print(np.shape(burndate_data))
print(qa_data.dtype)

# Clean up when done

del qa_data, qa_handle, burndate_data, burndate_handle

14
Example 3: MATLAB code to read the MCD64A1 Burn Date SDS using the MATLAB routine hdfread.
This is probably the easiest way to read individual HDF SDSs in MATLAB.

mcd64a1_file = ’MCD64A1.A2006213.h20v11.061.2021105210759.hdf’

% read burn date layer in one shot

burn_date = hdfread(mcd64a1_file, ’Burn Date’);

% display burn date (transposed so that North appears on top)

imagesc(burn_date.’);

Example 4: Longer version of MATLAB code to read the MCD64A1 Burn Date layer. This is probably the
better approach to use if multiple SDSs or multiple spatial subsets of an SDS will be read since the HDF
file will be opened and closed only once. (The shorter approach using hdfread requires that the file be
opened and closed for each read.)

mcd64a1_file = ’MCD64A1.A2006244.h31v10.061.2021106034604.hdf’

sd_id = hdfsd(’start’, mcd64a1_file, ’DFACC_RDONLY’);

sds_index = hdfsd(’nametoindex’, sd_id, ’Burn Date’);

sds_id = hdfsd(’select’, sd_id, sds_index);

% prepare to read entire SDS (always 2400 x 2400 pixels in size)

start = [0,0];
edges = [2400,2400];

[burn_date, status] = hdfsd(’readdata’, sds_id, start, [], edges);

status = hdfsd(’endaccess’, sds_id);

status = hdfsd(’end’, sd_id);

% display burn date (transposed so that North appears on top)

imagesc(burn_date.’);

15
Example 5: C code for reading MCD64A1 burn date layer using HDF library functions.
#include <stdio.h>
#include <stdlib.h>
#include "mfhdf.h"

#define ROWS 2400

#define COLS 2400

main(int argc, char **argv)

{
int32 sd_id, sds_index, sds_id;
int32 rank, data_type, nattr, dim_sizes[MAX_VAR_DIMS];
int32 start[2], int32 edges[2];

char *infile;
int i, j;
long nburn;
uint16 burn_date[ROWS][COLS];

infile = "MCD64A1.A2006244.h31v10.061.2021106034604.hdf";

if ((sd_id = SDstart(infile, DFACC_READ)) == FAIL) exit(1);

start[0] = start[1] = 0;
edges[0] = ROWS;
edges[1] = COLS;

if ((sds_index = SDnametoindex(sd_id, "Burn Date")) == FAIL) exit(2);

if ((sds_id = SDselect(sd_id, sds_index)) == FAIL) exit(3);
if (SDgetinfo(sds_id, (char *) NULL, &rank, dim_sizes, &data_type,
&nattr) == FAIL) exit(4);

/* check rank and data type */

if (rank != 2) exit(5);
if (data_type != DFNT_UINT16) exit(6);

if (SDreaddata(sds_id, start, NULL, edges,

(void *) burn_date) == FAIL) exit(7);

if (SDendaccess(sds_id) == FAIL) exit(8);

if (SDend(sd_id) == FAIL) exit(9);

/* simple example: count burned grid cells */

nburn = 0L;
for (i = 0; i < ROWS; i++) {
for (j = 0; j < COLS; j++)
if (burn_date[i][j] > 0) nburn++;
}

printf("%d burned grid cells.\n", nburn);

exit(0);
}

16
3.2 GeoTIFF subset for GIS visualization and analysis: MCD64monthly
A user-friendly GeoTIFF version of the MCD64 product is derived from the standard MCD64A1 HDF
version by University of Maryland. The GeoTIFF files are reprojected in Plate-Carrée projection and cover a
set of sub-continental windows (Figure 4). A table containing the regions covered and bounding coordinates
of the 24 windows is available in Appendix A.

-135 -90 -45 0 45 90 135

60

60
1 2 17
8 15 16
22
3
30

30
24
4 23 9 10 11 18
19
5
0

0
12
6 14
13 20
-30

-30
7 21
-60

-60
-135 -90 -45 0 45 90 135

Figure 4: Coverage of the GeoTIFF subsets. A table of bounding coordinates is available in Appendix A.

3.2.1 Naming Convention

The GeoTIFF files follow a naming convention similar to the official MCD64A1 product. However, as the
GeoTIFF files are obtained by mosaicing, resampling, and reprojecting several tiles of the original product,
the processing time is not available. Example product file names are:

MCD64monthly.A2000306.Win01.061.burndate.tif
MCD64monthly.A2000306.Win01.061.ba_qa.tif

where
MCD64monthly = monthly GeoTIFF version of MCD64A1
A2000306 = year and ordinal date of the starting day of the month covered by the product: 306 is the
ordinal date of Nov 1, hence 2000306 means that the product covers November 2000.
Win01 = spatial extent: the file covers window 1 (Alaska)
061 = version identifier (Collection 6.1)
burndate/ba qa = file content: unlike HDF, GeoTIFF files contain a single layer. Currently, two layers
of the original product are available as GeoTIFF files: Burn Date and QA.

17
3.2.2 Example Code
Example 6: IDL code to read the GeoTIFF MCD64monthly Burn Date product. Although not shown in this
example, the IDL QUERY TIFF function can be used to determine information about the GeoTIFF image
without having to read it into memory.

infile = ’MCD64monthly.A2016183.Win13.061.burndate.tif’

; read entire image

burn_date = read_tiff(infile, GEOTIFF=geo)

; now read just a spatial subset

burn_date_subset = read_tiff(infile, SUB_RECT = [1000, 1000, 400, 400])

help, burn_date
help, burn_date_subset
help, geo, /STRUCT

The code produces the following output:

BURN_DATE INT = Array[7055, 4552]
BURN_DATE_SUBSET
INT = Array[400, 400]
** Structure <18147a8>, 10 tags, length=120, data length=114, refs=1:
MODELPIXELSCALETAG
DOUBLE Array[3]
MODELTIEPOINTTAG
DOUBLE Array[6, 1]
GTMODELTYPEGEOKEY
INT 2
GTRASTERTYPEGEOKEY
INT 2
GTCITATIONGEOKEY
STRING ’Geographic (Longitude, Latitude) Unspe’...
GEOGRAPHICTYPEGEOKEY
INT 32767
GEOGGEODETICDATUMGEOKEY
INT 32767
GEOGANGULARUNITSGEOKEY
INT 9102
GEOGSEMIMAJORAXISGEOKEY
DOUBLE 0.0000000
GEOGSEMIMINORAXISGEOKEY
DOUBLE 0.0000000

18
3.3 Shapefile subset for GIS visualization and analysis: MCD64monthly
Shapefiles of the MCD64A1 Burn Date layer are derived from the monthly GeoTIFF files by the University
of Maryland. The shapefiles are available with the same projection (Plate-Carrée) and geographic extent
used for the GeoTIFF sub-continental windows (Figure 4).

3.3.1 Naming Convention

The shapefile naming convention is identical to the GeoTIFF naming convention. Each shapefile consists of
multiple files which must remain in the same subdirectory. For convenience, shapefiles are distributed as a
zipped tar archive file (“.tar.gz”) containing the four shapefile elements. MCD64monthly shapefile archives
are named as follows:

MCD64monthly.A2000306.Win01.061.burndate.shapefiles.tar.gz

which in turn contains the following files:

MCD64monthly.A2000306.Win01.061.burndate.shp
MCD64monthly.A2000306.Win01.061.burndate.shx
MCD64monthly.A2000306.Win01.061.burndate.prj
MCD64monthly.A2000306.Win01.061.burndate.dbf

where
MCD64monthly = monthly shapefile version of MCD64A1
A2000306 = year and ordinal date of the starting day of the calendar month covered by the product (here
November 2000).
Win01 = spatial extent: the file covers window 1 (Alaska)
061 = version identifier (Collection 6)
burndate = file content.

19
3.4 MCD64CMQ Climate Modeling Grid Burned Area Product
The CMG burned area product is a gridded summary intended for use in regional and global modeling.

3.4.1 Naming Convention

The MCD64CMQ monthly product files follow the standard MODIS product naming convention. Examples:

MCD64CMQ.A2003213.061.2022183041945.hdf August 2003 product file

MCD64CMQ.A2003214.061.2022183041945.hdf August 2004 product file

3.4.2 Global Grid

Each data layer has 720 rows and 1440 columns comprising a global 0.25◦ grid. The center of the upper left
grid cell is located at longitude 179.875◦ W, 89.875◦ N. One data layer has an additional dimension (i.e., it is
three dimensional) as described below.

3.4.3 Data Layers

Each product file nominally contains four data layers (BurnedArea, QA, UnmappedFraction, and Land-
CoverDist), each stored as a separate HDF4 Scientific Data Set (SDS). Note that the LandCoverDist SDS
will be absent from the most recent MCD64CMQ product files due to the ∼1-year forward production lag
in the MCD12Q1 annual land cover product used to populate this layer. Such incomplete MCD64CMQ
product files have a global string attribute named LandCoverNote and will be reprocessed when the relevant
annual MCD12Q1 product becomes available.

BurnedArea: Monthly area burned. The scaled values stored in this SDS must be multiplied by a factor of
0.01 to obtain burned area in hectares.

QA: 8-bit quality assurance bit field. 0 = water, 1 = unprocessed land, 2 = processed land.

UnmappedFraction: Combined spatial and temporal fraction of the unmapped land area within each grid
cell [percent]. Specifically, this quantity is the average number of days that could be mapped during
the one-month product period over all 500-m land grid cells falling within each 0.25◦ bin, divided
by the number of days in the product calendar month. The resulting proportion is expressed as a
percentage.
Note that the unmapped fraction does not merely tally those 500-m grid cells in the MCD64A1 product
classified as unmapped (such grid cells have a value of -1 in the MCD64A1 Burn Date SDS), for which
the number of days mapped is zero. Rather, the quantity represents an average over all 500-m land
grid cells, whether they are entirely unmapped, partially mapped due to a shortened mapping period
(bit 2 set in the MCD64A1 QA SDS), or fully mapped during the product period.
Unprocessed land grid cells (see QA layer above) will have an unmapped fraction of 100%.

LandCoverDist: A breakdown (to the nearest percent) of the area burned in each grid cell by the 16 dif-
ferent UMD land cover classes available in the Collection 6 MCD12Q1 land cover product. These
classes, which have been slightly modified for Collection 6, are shown below. The number in each
row is the index into the specific plane of the three-dimensional LandCoverDist array that contains
the percentage of the area burned for the corresponding land cover type.

20
 0 water 8 woody savannas
1 evergreen needleleaf forests 9 savannas
2 evergreen broadleaf forests 10 grasslands
3 deciduous needleleaf forests 11 permanent wetlands
4 deciduous broadleaf forests 12 croplands
5 mixed forests 13 urban and built-up
6 closed shrublands 15 barren
7 open shrublands 16 unclassified

Note: The Collection 6 MCD12Q1 UMD land cover layer suffers from a bug that often assigns 500-m
grid cells an invalid class value of 14 (the IGBP “cropland/natural vegetation mosaics” class).2 This
incorrect class propogates into the MCD64CMQ product, and we therefore strongly advise users to ig-
nore the LandCoverDist layer until the problem is corrected. In the interim all new C6 MCD64CMG
product files are being being generated without the LandCoverDist layer. We will regenerate the
MCD64CMQ product using the corrected Collection 6.1 MCD12Q1 product when it becomes avail-
able in 2022.

3.4.4 Metadata
File-level metadata are stored as global attributes in the HDF product files (Table 4).

Table 4: MCD64CMQ metadata stored as standard global HDF attributes.

Attribute Name Description

ShortName Short name of product (“MCD64CMQ”).
Instrument Instrument name (“MODIS”).
BinSize Global grid bin size in degrees (0.25).
StartDate Start date of monthly product to nearest second.
EndDate End date of monthly product to nearest second.
NumInputBA Number of MCD64A1 input files (nominally 268).
InputPointerBA Comma-separated list of MCD64A1 input file names.
NumInputLC Number of MCD12Q1 input files.
InputPointerLC Comma-separated list of MCD12Q1 input file names.
LandCoverNote String attribute present only in product files lacking the LandCoverDist SDS. Ex-
ample value: “2018 MCD12Q1 product not available at time of processing”.

2
See https://landweb.modaps.eosdis.nasa.gov/cgi-bin/QS/new/pages.cgi?name=known_issues,
case number PM MCD12 19074.

21
3.4.5 Example Code
Example 7: IDL code to read the MCD64CMQ BurnedArea SDS.

pro cmg_demo
COMPILE_OPT STRICTARR, STRICTARRSUBS

mcd64cmq_file = ’MCD64CMQ.A2018001.006.2018149165520.hdf’

sd_id = HDF_SD_START(mcd64cmq_file, /READ)

index = HDF_SD_NAMETOINDEX(sd_id, ’BurnedArea’)

sds_id = HDF_SD_SELECT(sd_id, index)

; read SDS scale factor attribute

attr_index = HDF_SD_ATTRFIND(sds_id, ’scale_factor’)
HDF_SD_ATTRINFO, sds_id, attr_index, DATA=scale_factor
; IDL returns scalar numeric attributes as a 1-element vector
; next line will convert it to a scalar
scale_factor = scale_factor[0]

; read SDS units attribute

attr_index = HDF_SD_ATTRFIND(sds_id, ’units’)
HDF_SD_ATTRINFO, sds_id, attr_index, DATA=units

; read scaled data

HDF_SD_GETDATA, sds_id, ScaledAburn
HDF_SD_ENDACCESS, sds_id

HDF_SD_END, sd_id

help, ScaledAburn, scale_factor, units

Aburn = scale_factor * temporary(ScaledAburn)

help, Aburn

; calculate total area burned worldwide

Atotal = total(Aburn, /DOUBLE) ; ha

print, Atotal/1.0e6, FORMAT=’("Total burned area is",F6.2," Mha")

end

The code produces the following output:

SCALEDABURN LONG = Array[1440, 720]

SCALE_FACTOR FLOAT = 0.0100000
UNITS STRING = ’hectares’
ABURN FLOAT = Array[1440, 720]
Total burned area is 26.91 Mha

22
Example 8: IDL code to generate vectors containing the the latitude and longitude at the center of each
0.25◦ grid cell.

pro cmg_nav
COMPILE_OPT STRICTARR, STRICTARRSUBS

XGRID = 1440
YGRID = 720
bin_size = 0.25 ; degrees

lat = (90.0 - bin_size/2) - bin_size * indgen(YGRID) ; degrees

lon = (-180.0 + bin_size/2) + bin_size * indgen(XGRID) ; degrees

; print center latitude and longitude of first 24 rows/columns of grid

help, lat
print, lat[0:23], FORMAT=’(8F9.3)’

help, lon
print, lon[0:23], FORMAT=’(8F9.3)’
end

The code produces the following output:

LAT FLOAT = Array[720]

89.875 89.625 89.375 89.125 88.875 88.625 88.375 88.125
87.875 87.625 87.375 87.125 86.875 86.625 86.375 86.125
85.875 85.625 85.375 85.125 84.875 84.625 84.375 84.125
LON FLOAT = Array[1440]
-179.875 -179.625 -179.375 -179.125 -178.875 -178.625 -178.375 -178.125
-177.875 -177.625 -177.375 -177.125 -176.875 -176.625 -176.375 -176.125
-175.875 -175.625 -175.375 -175.125 -174.875 -174.625 -174.375 -174.125

23
4 Obtaining the MODIS Burned Area Products
All MODIS products are available free of charge. The MODIS Burned Area Product is available for order-
ing from the Land Processes Distributed Active Archive Center (LP-DAAC).3 Additionally, a secure FTP
(SFTP) server is maintained by the University of Maryland, primarily to assist science users who need to
regularly download large volumes of data.

4.1 Downloading the products from the fuoco SFTP server

The MODIS burned area product is available for download in either HDF, GeoTIFF, or Shapefile format
from the University of Maryland fuoco SFTP (formerly FTP) server. Connect using the following infor-
mation:

Server: fuoco.geog.umd.edu
Login name: fire
Password: burnt

Note: As a consequence of our mandatory transition from FTP to SFTP4 , most users will not be able to
download product files using their regular web browser without first installing a third-party browser exten-
sion.

For downloading product files you can use the command-line sftp and lftp clients, or freely available
GUI file transfer software such as FileZilla (https://filezilla-project.org) and Cyberduck
(https://cyberduck.io/). SFTP-capable commercial software is also available and includes the ex-
amples listed below.

For Windows:

• SmartFTP (https://www.smartftp.com/)

• WinSCP (https://winscp.net)

For MacOS:

• ForkLift (https://binarynights.com/)

• Commander One (https://mac.eltima.com/file-manager.html)

• Transmit (https://panic.com/transmit/)

• Viper FTP (https://viperftp.com/)

• Flow (http://fivedetails.com/)

3
https://lpdaac.usgs.gov/
4
In our case the change in protocol was not driven by a need for encryption – here such a requirement would be pointless given
that the data are intended to be freely available and we openly share the password – but instead to avoid various port-level server
vulnerabilities associated with FTP.

24
Once connected, enter the data subdirectory to access the following abridged directory tree:

.
|-- GFED
|-- MODIS
| |-- C6
| ‘-- C61
| |-- docs
| |-- MCD64A1
| | |-- HDF
| | | ‘-- h00v08, h00v09, ..., h35v10
| | |-- SHP
| | | ‘-- Win01, Win02, ..., Win24
| | | ‘-- 2000, 2001, 2002, ...
| | ‘-- TIFF
| | ‘-- Win01, Win02, ..., Win24
| | ‘-- 2000, 2001, 2002, ...
| ‘-- MCD64CMQ
|-- VIIRS
‘-- VIRS

Note: The SHP and TIFF subdirectories are currently empty but will be populated later this year.

4.1.1 HDF Files

As can be seen from the directory tree, the file system on this server is structured to organize the HDF
MCD64A1 product hierarchically by tile. Specifically, all MCD64A1 HDF product files for a given tile are
located in the directory MODIS/C6/MCD64A1/HDF/hHHvVV, where HH is the horizontal tile coordinate
and VV is the vertical tile coordinate. For example, the Collection 6 MCD64A1 HDF files for MODIS tile
h08v05 will be found in the directory MODIS/C61/MCD64A1/HDF/h08v05.

4.1.2 GeoTIFF files and Shapefiles

The layout of GeoTIFF files and Shapefiles on the server is structured to organize the data hierarchically by
window, and then by year. All the data for the same window from the same year is located (for GeoTIFF
files and shapefiles, respectively) in a the directories MODIS/C61/MCD64A1/TIF/WinXX/YYYY/ and
MODIS/C61/MCD64A1/SHP/WinXX/YYYY/, where XX is the number of the window (Figure 4) and
YYYY is the year. For example, the directory MODIS/C61/MCD64A1/Win01/2001 contains all of the
GeoTIFF files for the year 2001 for window 01 (Alaska).

25
4.1.3 Example sftp command line session
Here we use wildcards with the SFTP get command to download the January 2013 MCD64A1 HDF prod-
uct for the entire globe (i.e., all 268 MODIS land tiles for which the product is produced).
$ sftp [email protected]
Password:
Connected to fuoco.geog.umd.edu.
sftp> cd data
sftp> ls
GFED MODIS VIIRS VIRS
sftp> cd MODIS/C61/MCD64A1/HDF
sftp> progress
Progress meter disabled
sftp> get h??v??/MCD64A1.A2013001*
Fetching /data/MODIS/C61/MCD64A1/HDF/h00v08/MCD64A1.A2013001.h00v08.061.2021...
to MCD64A1.A2013001.h00v08.061.2021013041920.hdf
Fetching /data/MODIS/C61/MCD64A1/HDF/h00v09/MCD64A1.A2013001.h00v09.061.2021...
to MCD64A1.A2013001.h00v09.061.2021013052550.hdf
Fetching /data/MODIS/C61/MCD64A1/HDF/h00v10/MCD64A1.A2013001.h00v10.061.2021...
to MCD64A1.A2013001.h00v10.061.2021013041341.hdf
.
.
.
Fetching /data/MODIS/C61/MCD64A1/HDF/h35v08/MCD64A1.A2013001.h35v08.061.2021...
to MCD64A1.A2013001.h35v08.061.2021013053444.hdf
Fetching /data/MODIS/C61/MCD64A1/HDF/h35v09/MCD64A1.A2013001.h35v09.061.2021...
to MCD64A1.A2013001.h35v09.061.2021013053444.hdf
Fetching /data/MODIS/C61/MCD64A1/HDF/h35v10/MCD64A1.A2013001.h35v10.061.2021...
to MCD64A1.A2013001.h35v10.061.2021013042434.hdf
sftp> bye

26
5 Working with the product in ENVI 5.5
Provided here are instructions for opening the MCD64A1 HDF, GeoTIFF, and Shapefile products for both
the standard and Classic ENVI 5.5 interfaces.

5.1 MCD64A1 (HDF)

HDF MODIS products are only partially supported in ENVI. To open them, select File → Open As Scientific
Formats → HDF4. In the Dataset Browser (Figure 5), load the Data Field (e.g., Burn Date) into the Raster
field (center top button). Opened as a generic HDF file, all geographic information is lost. To restore this
information the projection parameters must be entered manually.

Figure 5: ENVI 5.5 example Dataset Browser.

ENVI Classic
HDF MODIS products are only partially supported in ENVI Classic. To open them, select File → Open
External Files → Generic Formats → HDF. Select the layers of interest , which are labeled by their HDF4
long-name attribute, e.g., “ordinal day of burn” (Figure 6). Opened as a generic HDF file, all geographic
information is lost. To restore this information the projection parameters must be entered manually.

27
 Figure 6: ENVI 5.5 Classic example Dataset Browser and image display.

5.2 MCD64monthly (GeoTIFF)

The GeoTIFF product files are fully supported in ENVI. Use the File → Open... menu to open them.

ENVI Classic
The GeoTIFF product files are fully supported in ENVI Classic. Use the File → Open Image File menu to
open them.

5.3 MCD64monthly (Shapefile)

Shapefiles are fully supported in ENVI (but not ENVI Classic – see below). To open a Shapefile, simply go
through the File → Open... menu. To view metadata or attributes, right-click on the file name to open the
sub-menu (Figure 7). To change the vector colors, select Properties on the sub-menu.

ENVI Classic
Shapefiles are not directly supported in ENVI Classic, rather they are converted to the ENVI Vector File
(“.evf”) format during ingest. To load a shapefile, select Vector → Open Vector File and choose the Shapefile
wildcard filter setting (“*.shp”). When prompted to set “Import Vector Files Parameters”, set the desired

28
Figure 7: ENVI 5.5 Shapefile (colored polygons) with the available sub-menu and vector properties window.
The different colored polygons represent the individual day of burn. Other color options are available.

layer name and output file location and select “OK”. Do not adjust the projection information; the default
values correspond to the Plate-Carrée projection used by the GeoTIFF files. Use the “Available Vectors List”
(Vector → Available Vectors List) to overlay the vectors on an existing display, or display them in a new
window.

29
6 Working with the product in ArcGIS
Provided here are instructions for opening the MCD64A1 HDF, GeoTIFF, and Shapefile products in ArcGIS
Pro 2.6.1 and ArcGIS ArcMap Desktop 10.8.1.

6.1 MCD64A1 (HDF)

6.1.1 ArcGIS Pro
HDF-EOS files are fully supported in ArcGIS Pro. To open them, simply go through the Add Data → Mul-
tidimensional Raster Data menu to import the desired HDF file(s) (Figure 8) and select the desired layer(s)
to import (Figure 9). Geographic information is automatically applied to the data.

Figure 8: Importing HDF data into ArcGIS Pro.

30
 Figure 9: Selecting the HDF Burn Date layer in ArcGIS Pro.

In order to display the burned areas only (Figure 10), go to Symbology → Unique Values and change the
following values to “No Color”:

• 0 - unburned

• -1 - missing data

• -2 - water

ArcGIS Pro will display the burn date as the ordinal day of the product year in either individual colors
or in the same color depending on your settings.

31
 Figure 10: January 2021 MCD64A1 Burn Date layer for MODIS tile h16v07 displayed in ArcGIS Pro.

6.1.2 ArcGIS ArcMap Desktop

HDF-EOS files are fully supported in ArcMap. To open them, simply go through the Map → Add Data
menu and select the desired layers to import (e.g., Burn Date). Geographic information is automatically
applied to the data.

6.2 MCD64monthly (GeoTIFF)

The MCD64 GeoTIFF files can be directly loaded into ArcGIS through Add Data. In order to display the
burned areas only, go to Symbology → Unique Values and change the following values to “No Color”:
• 0 - unburned
• -1 - missing data
• -2 - water
ArcGIS will display the burn date as the ordinal day of the product year in either individual colors or in
the same color depending on your settings.

6.2.1 Area of Interest (AoI)

In order to reduce the file size of the regional GeoTIFF and focus on a specific region, an “area of interest”
(AoI) spatial subset can be extracted.

32
ArcGIS Pro

1. Display desired AoI

2. Highlight the TIFF layer by clicking its name in the Contents pane

3. Select Data → Export Raster → Clipping Geometry → Current Display Extent → Export Data

The “Export Raster Data” window will open (Figure 11).

Figure 11: ArcGIS Pro export raster window.

ArcGIS ArcMap Desktop

1. Display desired AoI

2. Right mouse-click the data layer

3. Select Data → Export Data

The “Export Raster Data” window will open (Figure 12). Check both data frames options to Current and
choose TIFF in the Format field.

6.3 MCD64monthly (Shapefile)

The Shapefiles can be loaded directly into ArcGIS and ArcMap via Add Data. To change the appearance of
the file, select Layer Properties → Symbology → Quantities → Graduated Colors. Optionally, to remove
the outlines of the data, right click a symbol in the Graduated Colors panel and select “Properties For All
Symbols”, then set “Outline Color” to “No Color”. An example is shown in Figure 13.

33
 Figure 12: ArcGIS ArcMap export raster window.

Figure 13: Left: True-color (R, G, B) Sentinel-2 MSI scene acquired 19 September 2019 in Saskatchewan,
Canada. Right: MCD64monthly September Shapefile superimposed over the burned area. Older burned
areas are shown in blue tones, while newer burns are shown in red tones.

34
7 Validation of the MODIS Burned Area Product
The Collection 6 product has been validated at Stage 3 by comparison with Landsat 8 Operational Land
Imager (OLI) image pairs acquired 16 days apart that were visually interpreted. These independent refer-
ence data were selected using a stratified random sampling approach that allows for probability sampling of
Landsat data in both time and in space. A total of 558 Landsat 8 OLI image pairs (1116 images), acquired
between 1 March 2014 and 19 March 2015, were selected. Globally, the MCD64A1 product had an esti-
mated 40.2% commission error and 72.6% omission error; the prevalence of omission errors is reflected by a
negative estimated bias of the mapped global area burned relative to the Landsat independent reference data
(-54.1%). Globally, the standard errors of the accuracy metrics were less than 6%. The product accuracy
was also characterized at coarser scale using metrics derived from the regression between the proportion of
coarse resolution grid cells detected as burned by MCD64A1 and the proportion mapped in the Landsat 8
interpreted maps. The errors of omission and commission observed at 30-m resolution compensate to a
considerable extent at coarser resolution, as indicated by the coefficient of determination (r2 > 0.70), slope
(> 0.79) and intercept (-0.0030) of the regression between the MCD64A1 product and the Landsat indepen-
dent reference data in 3-km, 4-km, 5-km, and 6-km coarse resolution cells. The analysis of the deviations
between the proportion of area burned mapped by the MCD64A1 product and by the independent reference
data, performed using 3-km × 3-km and 6-km × 6-km coarse resolution cells, indicates that the large neg-
ative bias in global area burned is primarily due to the systematic underestimation of smaller burned areas
in the MCD64A1 product. Full details of the product validation may be found in Bochetti et al. (2016) and
Boschetti et al. (2019).

8 Known Problems
8.1 Cropland Burning
Burned areas in cropland should generally be treated as low confidence due to the inherent difficulty in
mapping agricultural burning reliably. See Hall et al. (2016) for more information about the complications
unique to cropland-fire mapping.

8.2 Pre-November 2000 Mission Period

Although the Terra MODIS first began acquiring data in February 2000, crosstalk and calibration remained
problematic until early November 2000. For this reason we do not distribute the MODIS burned area product
for the mission period preceding November 2000.

8.3 June 2001 Terra MODIS Outage

The June 2001 monthly product is heavily degraded due to extended Terra MODIS outages.

35
8.4 August 2020 Aqua Outage
A failure of the Aqua formatter-multiplexer unit (FMU) on 16 August 2020 led to the loss of regular science
data telemetry for a period of about two weeks. During this time the MODIS instrument remained otherwise
functional, and an effort was made to generate the standard Aqua science products from Direct Broadcast
Aqua MODIS data collected by Direct Readout ground stations around the world. A representative example
of the limited Direct Broadcast coverage obtained during the outage is shown in in Figure 14.
While essentially all Aqua science data over Africa, eastern Asia, Indonesia, and Oceana were lost
during the two-week outage period, we do not expect the loss to significantly degrade the MCD64A1 burned
area product in these regions since the Terra MODIS continued to function normally.

Figure 14: Representative Direct Broadcast coverage available during the 16 August – 2 September 2020
Aqua FMU outage period, using the MODIS active fire (not burned area) product as an example. Fire pixels
are shown in red, clear land pixels are shown in green, cloud pixels are shown in light blue, and areas lacking
data are shown in white. Browse image courtesy of the LDOPE.

9 Frequently Asked Questions

Is there an existing tool I can use to remap the tiled MODIS products into a different projec-
tion?
The now-defunct MODIS Reprojection Tool (MRT) and follow-on HDF-EOS to GeoTIFF Conversion Tool
(HEG) can remap the tiled MODIS products into many different projections; see Section 11.

36
How do I calculate the latitude and longitude of a grid cell in the Level 3 products?
You can use the online MODLAND Tile Calculator5 , or perform the calculation as described in Appendix B.

How do I calculate the tile and grid cell coordinates of a specific geographic location (latitude
and longitude)?
You can use the online MODLAND Tile Calculator, or perform the calculation as described in Appendix B.

10 References
Boschetti, L., Roy, D. P., Giglio, D. P., Huang, H., Zubkova, M., and Humber, M. L., 2019, Global validation
of the collection 6 MODIS burned area product. Remote Sensing of Environment, 235, 11490.

Boschetti, L., Stehman, S. V., and Roy, D. P., 2016, A stratified random sampling design in space and time
for regional to global scale burned area product validation. Remote sensing of environment, 186, 465-478.

Giglio, L., Boschetti, L., Roy, D. P., Humber, M. L., and Justice, C. O., 2018, The Collection 6 MODIS
burned area mapping algorithm and product. Remote Sensing of Environment, 217, 72-85.

Giglio, L., Loboda, T., Roy, D. P., Quayle, B., and Justice, C. O., 2009, An active-fire based burned area
mapping algorithm for the MODIS sensor. Remote Sensing of Environment, 113, 408-420.

Hall, J. V., Argueta, F., and Giglio, L., 2021, Validation of MCD64A1 and FireCCI51 cropland burned area
mapping in Ukraine. International Journal of Applied Earth Observation and Geoinformation, 102, 102443.

Hall, J. V., Loboda, T. V., Giglio, L., and McCarty, G. W., 2016, A MODIS-based burned area assessment
for Russian croplands: Mapping requirements and challenges. Remote Sensing of Environment, 184, 506-
521.

Humber, M., Boschetti, L., and Giglio, L., 2020, Assessing the shape accuracy of coarse resolution burned
area identifications. IEEE Transactions on Geoscience and Remote Sensing, 58, 15161525.

Roy, D. P., Giglio, L., Kendall, J. D., and Justice, C. O., 1999, A multitemporal active-fire based burn scar
detection algorithm. International Journal of Remote Sensing, 20, 1031-1038.

5
http://landweb.nascom.nasa.gov/cgi-bin/developer/tilemap.cgi

37
11 Relevant Web and FTP Sites
• MODIS Fire and Thermal Anomalies: General information about the MODIS Fire (Thermal Anoma-
lies) and Burned Area products.
https://modis-fire.umd.edu/

• MODIS Land: Information about the MODIS Land products.

https://modis-land.gsfc.nasa.gov/index.html

• MODIS Land Product Quality Assessment: Product quality-assessment (QA) related information,
including a very complete archive of known land-product issues with descriptions and examples.
https://landweb.modaps.eosdis.nasa.gov/cgi-bin/QS/new/index.cgi

• HEG: HDF-EOS to GeoTIFF Conversion Tool: Software for reprojecting tiled HDF-EOS products
into many different projections.
https://wiki.earthdata.nasa.gov/display/DAS/HDF-EOS2+to+HDF-EOS5+Conversion+Tool

• MODLAND Tile Calculator: Online tool for performing forward and inverse mapping of MODIS
sinusoidal tiles.
http://landweb.nascom.nasa.gov/cgi-bin/developer/tilemap.cgi

• NASA Earthdata Search: The primary distribution site for most MODIS products. Formerly Reverb,
and before that the Warehouse Inventory Search Tool (WIST), and before that the EOS Data Gateway
(EDG).
https://search.earthdata.nasa.gov/

38
Appendix A Coverage of the GeoTIFF subsets

Table 5: Regions and bounding coordinates of the GeoTIFF subsets.

Min. Max. Min. Max.

Window Coverage Lon. Lon. Lat. Lat.
1 Alaska -180 -140.5 50 70
2 Canada -141 -50 40 70
3 USA (Conterminous) -125 -65 23 50
4 Central America -118 -58 7 33
5 South America (North) -82 -34 -10 13
6 South America (Central) -79 -34 -35 -10
7 South America (South) -77 -54 -56 -35
8 Europe -11 35 33 70
9 West and North Africa -19 5 0 37.5
10 Central and North Africa 5 25 0 37.5
11 East Africa and Arabian Peninsula 25 65 0 37.5
12 Southern Africa (North) 8.5 48 -15 5.5
13 Southern Africa (South) 10 41 -35 -15
14 Madagascar 42 59 -27 -10
15 Russia and Central Asia 1 35 90 33 70
16 Russia and Central Asia 2 90 145 33 70
17 Russia (Kamachatka) 145 180 40 70
18 South Asia 60 93 5 36
19 South East Asia 90 155 -10 33
20 Australia 112 155 -45 -10
21 New Zealand 165 179 -48 -33
22 Azores -31.6 -24.8 36.8 40
23 Cape Verde Island -25.5 -22.5 14.6 17.5
24 Hawaii -161 -154 18 24

39
Appendix B Coordinate conversion for the MODIS sinusoidal projection
Navigation of the tiled MODIS products in the sinusoidal projection can be performed using the forward
and inverse mapping transformations described here. We’ll first need to define a few constants:

R = 6371007.181 m, the radius of the idealized sphere representing the Earth;

T = 1111950 m, the height and width of each MODIS tile in the projection plane;

xmin = -20015109 m, the western limit of the projection plane;

ymax = 10007555 m, the northern limit of the projection plane;

w = T /2400 = 463.31271653 m, the actual size of a “500-m” MODIS sinusoidal grid cell.

B.1 Forward Mapping

Denote the latitude and longitude of the location (in radians) as φ and λ, respectively. First compute the
position of the point on the global sinusoidal grid:

x = Rλ cos φ (1)
y = Rφ. (2)

Next compute the horizontal (H) and vertical (V ) tile coordinates, where 0 ≤ H ≤ 35 and 0 ≤ V ≤ 17
(Section 1.2.2):
x − xmin
 
H = (3)
T
ymax − y
 
V = , (4)
T

where ⌊ ⌋ is the floor function. Finally, compute the row (i) and column (j) coordinates of the grid cell
within the MODIS tile:
(ymax − y) mod T
 
i = (5)
w
(x − xmin ) mod T
 
j = . (6)
w
Note that for all 500-m MODIS products on the sinusoidal grid 0 ≤ i ≤ 2399 and 0 ≤ j ≤ 2399.

B.2 Inverse Mapping

Here we are given the row (i) and column (j) in MODIS tile H, V . First compute the position of the center
of the grid cell on the global sinusoidal grid:

x = (j + 0.5)w + HT + xmin (7)

y = ymax − (i + 0.5)w − V T (8)

40
Next compute the latitude φ and longitude λ at the center of the grid cell (in radians):
y
φ = (9)
R
x
λ = . (10)
R cos φ

B.3 Applicability to 250-m and 1-km MODIS Products

With the following minor changes the above formulas are also applicable to the higher resolution 250-m and
500-m MODIS tiled sinusoidal products.

250-m grid: Set w = T /4800 = 231.65635826 m, the actual size of a “250-m” MODIS sinusoidal grid cell.
For 250-m grid cells 0 ≤ i ≤ 4799 and 0 ≤ j ≤ 4799.

1-km grid: Set w = T /1200 = 926.62543305 m, the actual size of a “1-km” MODIS sinusoidal grid cell.
For 1-km grid cells 0 ≤ i ≤ 1199 and 0 ≤ j ≤ 1199.

41