DATA FUSION FOR ENVIRONMENTAL ASSESSMENT: INTEGRATING MODIS
IMAGERY AND NEXRAD WEATHER RADAR
R. P. Watson ®, L. A. Scuderi
* K. K. Benedict”, T. R. Kuntz*
? CREATE - Center for Rapid Environmental Assessment and Terrain Evaluation, University of New Mexico,
Albuquerque, NM USA - (r
watson, tree, tera) (Qunm.edu
^ Earth Data Analysis Center, University of New Mexico, Albuquerque, NM USA — (kbene) @edac.unm.edu
KEY WORDS: Environment, Vegetation, Precipitation, Change Detection, Monitoring, Fusion, MODIS, Imagery
ABSTRACT:
This paper discusses the fusion of terrestrial NEXRAD weather radar with MODIS hyperspectral imagery. Issues of data fusion
relating to the differing character of terrestrial weather radar and hyperspectral satellite imagery are evaluated and discussed. The
integration of these disparate data types provides challenges particularly in the geometric registration of the radial pattern of weather
radar to the gridded imagery data of MODIS. Weather radar data are systematically collected over virtually all of the Continental
United States and provide spatially continuous and calibrated rainfall estimates at a nominal spatial resolution of 4 km. Several
MODIS Level 4 products (MOD 13 - Gridded Vegetation Indices [NDVI & EVI]; MOD 15 - Leaf Area Index [LAI] and Fraction of
Photosynthetically Active Radiation [FPAR]; MOD 17 Net Photosynthesis [PSN] and Net Primary Production [NPP]) provide
global validated sources of vegetation conditions at 1 km resolution. The short-term response of vegetation to varying local weather
conditions is difficult to assess due to the spatially discontinuous nature of precipitation. Understanding the vegetation response to
local weather events is particularly important in arid and semi-arid regions such as the American Southwest that are susceptible to
periodic droughts characterized by a high spatial variability in precipitation. The high temporal frequency of these data (hourly to
daily for NEXRAD, and one to eight days for MODIS) allow the analysis of short-term local vegetation response and ecosystem
change to be undertaken at a finer temporal scale than has previously been possible.
1. INTRODUCTION
1.1 Background
This paper discusses research on the relationship between two
modes of environmental remote sensing; terrestrial NEXRAD
weather radar and the Terra and Aqua based MODIS
hyperspectral imagery for the purposes of monitoring the short-
term response of grasslands to localized rainfall. The short-term
response of vegetation to varying local weather conditions is
difficult to assess due to the spatially discontinuous nature of
precipitation. Understanding the vegetation response to local
weather events is particularly important in arid and semi-arid
regions such as the American Southwest that are susceptible to
periodic droughts characterized by a high spatial variability in
precipitation.
The high temporal frequency of these data (hourly to daily for
NEXRAD, and one to eight days for MODIS) allow the
analysis of short-term local vegetation response and ecosystem
change to be undertaken at a finer temporal scale than has
previously been possible. The realization of this potential
however requires the integration of these fundamentally
different data sources.
1.2 Project Area
The study area covers a portion of the semi-arid eastern plains
of New Mexico, USA (Figure 1). This region consists primarily
of low relief grasslands with average elevations across the study
area ranging from approximately 1500 to 2500 meters. The
major agricultural/economic activity of this area includes dairy
farming and beef cattle production and is part of the region
known as “Cattle Feeding Country” where more than 7 million
fed cattle are marketed annually, representing fully 30 percent
of the fed cattle production in the United States. These
rangelands are a vital national resource where monitoring is
critical to efficient, and sustainable use.
{ Te
Northeastern As Lot
1 T—— E 8
New Mexico | - "és
Legend |
ee AM Pats Want quw UR P]
MEL Varie Ariel 4 i |
Grassland Code Al J | 125 : za : me. Kkornetens
Neer grassland T T T
E den | Study Area for Grassland Coverage | atem tt
Figure 1: Project Area
. . age . . {
Annual moisture availability is dominated by summé
monsoonal flow into the region from the Gulf of Mex?
Additional cold season frontal precipitation 1$ typicall
associated with low-pressure disturbances that track castwat
across the state producing moderate precipitation totals dure
the winter and early spring. Maximal precipitation occurs ?
906
Inter
—
July
temp
cont!
ther
The:
resp:
mois
Sens
conc
This
influ
has |
prod
2.1
The
aval
algo
Proc
thro
Prec
Env
3.1.)
the
Are:
proy
Pho
veg
eact
al,
FPA
info
plan
indi
gras
Fou
fron
thro
com
repr
day
to c
imag
valu
NE?
2.1;
coll
prov
à no
2.1.
con
use
natn
(Nat
Con:
base
3 stc
