Satellite Data Access – Giovanni, LAADS, and NEO Training Workshop in Partnership with BAAQMD Santa Clara, CA September 10 – 12, 2013 Applied Remote SEnsing.

Satellite Data Access – Giovanni, LAADS, and NEO

Training Workshop in Partnership with BAAQMD

Santa Clara, CASeptember 10 – 12, 2013

Applied Remote SEnsing Training (ARSET) – Air Quality

A project of NASA Applied Sciences1

• Giovanni provides access and simple analysis tools for Level 3 data (i.e., aggregated, smoothed, etc.)

• Most data at 1 degree resolution• No advanced programming is needed• A lot of pre-processing / assumptions

are behind the numbers, no quality flags

• Good for – Data availability check– Hypothesis generation– Big picture thinking

• LAADS is a data access portal for Level 2 MODIS data

• Select and customize raw data files in HDF format

• Customization delays data delivery• Data comes at original spatial

resolution and quality flags• Need programming software (Matlab,

IDL, etc.) to process the downloaded data

Raw data list

4

Customized file list

5

• NEO provides pseudo Level 2 data – processed for a regular fixed global grid at a resolution near the original Level 2 data

• Provide information in ArcGIS compatible raster format (GeoTIFF)

• Also in spreadsheet format• Currently does not support batch

downloading• Good to check on episodes

Example: The 2003 European heat wave

• Extremely hot: mean summertime temperature > 5 SDs of the 1961-1990 mean

• Extremely costly: billions of dollars lost in agriculture due to heat and drought

• Extremely dangerous: > 35,000 excess deaths

Background Paper

Study Objectives

• Describe the 2003 heat wave–Vegetation, surface temperature, and

energy balance• Compare this description with RCM

simulations• Examine the spatial variation of the

heat wave• Examine the impact of land cover type

on air temperature

RS data

Major Findings• Vegetation was severely affected across the

area, especially across central France• Surface temperature anomaly peaked at 15.4

°C.• At a finer scale, vegetation and surface

temperature anomalies were greater for crops and pastures than for forests

• Surface sensible heat flux enhanced by 48-61% during August

• Observations during this period agree with RCM predictions.

• There is potentially a soil moisture feedback on surface energy balance that makes the drought longer and worse.

Land Surface Temperature

LST is the temperature of the interface between the Earth’s surface and its atmosphere.

Land surface air temperature (LSAT) is the temperature of the air near the Earth´s surface which is routinely measured at 1.5 to 2 m.

Access MODIS Global LST data

• http://neo.sci.gsfc.nasa.gov/Search.html• Duration – week (actually 8 days)• Date range: Aug 5 - 13, 2001 and 2003• Datasets:

– Land Surface Temperature [Day] (8 day - Terra/MODIS)

– Land Surface Temperature Anomaly [Day] (8 day)

• LST anomaly: show daytime land surface temperature anomalies for a given day compared to the average conditions during that period between 2001-2010

Data analysis

• Download data format: KML for Google Earth, GeoTIFF floating point for ArcGIS

• Check out the link “about this dataset”

• Rename files for clarity• To open in GIS, change file

extension to .tif• Use the value ranges in the NEO

color table to stretch your plots in GIS

MODIS LST data analysis questions

• Examine and describe the spatial patterns of LST around Paris–Given the spatial resolution, can you

tell which part of Paris region has high/low LST?

• Examine and describe the differences between LSTs in 2001 and 2003.–Which year has higher LST?–How are the differences spatially

distributed?

MODIS LST data analysis questions

• Examine the spatial patterns of LST anomalies around Paris– Is 2001/2003 a hotter than normal

year?–How are the anomaly values spatially

distributed? i.e., are there areas with particularly high anomalies?

Normalized Difference Vegetation Index

NDVI• Green leaves have a reflectance < 20% in the 0.5

to 0.7 mm range (VIS) and ~ 60% in the 0.7 to 1.3 mm range (NIR).

• Dead or stressed vegetation reflects more red and less NIR.

• Negative values of NDVI (approaching -1): deep water.

• Values close to zero (-0.1 to 0.1): barren land or snow.

• Low positive values (~0.2 to 0.4): shrub and grassland

• high values (approaching 1): temperate and tropical rainforests. 

• The typical range is between about -0.1 (for a not very green area) to 0.6 (for a very green area). 

NDVI data analysis questions

• Examine and describe the spatial patterns of NDVI around Paris–Given the spatial resolution, what can

you say about the vegetation cover in this region?

• Examine and describe the differences between NDVIs in 2001 and 2003.–Which year has more stressed

vegetation?–How are the differences spatially

distributed?