As urbanization continues to grow, retrieval of aerosols in higher urbanized areas becomes more important and successful unbiased
retrievals in urban areas will become more important both for air quality and in climate applications. However, retrieval of AOD by
satellite remote sensing measurements over land is complicated by the fact that the Top of Atmosphere (TOA) reflectance is a
combination of the desired atmospheric path reflectance as well as the ground reflectance. To avoid this problem, AOD retrieval with
the MODIS instrument attempts to isolate “dark” pixels such as dense dark vegetation. To account for land surface reflection
properties, the latest MODIS retrieval algorithm over land tries to improve on surface albedo modeling using collocated MODIS and AERONET sky radiometer data to improve the VIS - SWIR ratios. However, the matchup data taken on a global scale still was heavily concentrated over vegetated areas since there are very few urban AERONET sites and the resulting surface models are incapable of describing the urban surface reflection ratios properly. In addition, as we move to higher spatial resolution such as Collect 6 3km retrievals, the ability to isolate dark pixels is significantly reduced and the ability to independently characterize the surface using MODIS land classification data is crucial to avoid biases and improve on retrievals. The purpose of this presentation is to demonstrate that the C005 surface models trained on a global scale and used by the MODIS algorithm to estimate the ground reflectance is not appropriate for urban areas and that the use of Urban based surface models together with a land classification flag can remove biases and improve retrieval.
In the standard MODIS Collection 5 aerosol retrieval algorithm, estimates of the surface albedo between the VIS
and MIR channels must be made. Unfortunately, the operational model used is not suitable for urban areas and
efforts to modify the required VIS-MIR surface spectral ratios for urban areas are needed to remove aerosol retrieval
biases. To address these issues, we use results based on the ASRVN product to provide regionally tuned surface
reflection ratios Using these values removes retrieval bias and improves resolution to 1.5 km. In addition, we note
explore the relationships for multiple urban sites and illustrate a general correspondence between the surface
reflection ratiosn and biases in AOD retrieval. Further validations of the surface reflection differences in urban
areas are illustrated using high resolution LANDSAT 7 imagery for vegetation / urban boundaries.
A combination of CIMEL radiometer and MODIS measurements are used to correct surface albedo models. In particular,
we show through an analysis of hyperspectral high resolution Hyperion data that the correlation coefficient assumption
underestimates ground albedo resulting in an overestimate of the VIS optical depth and operational collect 5 surface
model shows an incorrect trend between the MVI index and the surface correlations. Preliminary radiative transfer
calculations based on the same model show that this mechanism can help explain the observed overestimation and the
corrected models have been implemented for NYC and Mexico City with significantly improved AOD.
A combination of CIMEL radiometer and MODIS measurements are used to correct surface albedo models. In
particular, we show through an analysis of hyperspectral high resolution Hyperion data that the correlation coefficient
assumption underestimates ground albedo resulting in an overestimate of the VIS optical depth and operational collect 5
surface model shows an incorrect trend between the MVI index and the surface correlations. Preliminary radiative
transfer calculations based on the same model show that this mechanism can help explain the observed overestimation
and the corrected models have been implemented for NYC and Mexico City with significantly improved AOD.
In this paper, we focus on the use of simultaneous MODIS and AERONET sky radiometer data to refine
the surface albedo models regionally and improve on the current AOD operational retrieval. In particular,
over New York City, we show that the correlation coefficient assumption used in the MODIS Collection
(5) model between the VIS and MIR channels used for surface reflection parameterization are still severely
underestimated in comparison with high spatial imagery data from Hyperion thereby leading to an
underestimate in the VIS ground albedos and explaining the subsequent overestimate of the VIS optical
depth. Furthermore, we find that the VIS/MIR ratios depend only weakly on the scattering geometry
allowing us to generate a regional VIS/MIR surface reflectance correlation coefficient map at spatial
resolutions down to 1.5km. When applying the new VIS/MIR surface reflectance ratio model, we show the
MODIS and AERONET derived optical thickness agreement is significantly improved for the operational
10km resolution product. Moreover, we also show the high resolution surface model allows us to improve
the resolution of the retrieved AOD to 3km. Although direct comparisons for a given day can only be made
at the AERONET site, we find the AOD spatial variability from the improved MODIS retrievals is in far
better agreement with temporal statistics seen in the AERONET time series retrievals. In addition to that,
we also process and validate with another urban area, Mexico City, and the result is also significantly
improved by using refined regional VIS/MIR surface reflectance ratio model.
The Moderate Resolution Imaging Spectroradiometer (MODIS) and the Multi-angle Imaging Spectroradiometer (MISR) are two of the instruments aboard the Terra Earth Observing System (EOS). Algorithms for the retrieval of cloud-top heights have been implemented in order to get a product that can be applied in climate change studies, climate modeling and atmospheric research. Cloud height information can be used to analyze the Caribbean climate and to understand deforestation patterns on rain forests. The algorithms to retrieve this kind of information are based on CO2 slicing method and stereo matching methods. Cloud height information appears in terms of cloud top pressures. To compare MODIS cloud top pressures with MISR cloud top heights, it is important to look for a good atmospheric profile for the Caribbean such as by looking at field instrument observations. Available data from MODIS and MISR is geolocated in different latitudes and longitudes. MISR technique is an innovative method that assigns height values in a geometric form. In order to compare MODIS cloud-top pressures and MISR cloud-top heights, cloud-top pressures must be converted into cloud-top heights. Upper air observations can be used to get pressure-height profiles over the Caribbean. Also this kind of data can be used to validate MODIS and MISR parameters. Do cloud height measurements from MODIS can be better comparing to MISR measurements? Do cloud height measurements from MODIS or MISR can be used to classify cloud types? How confident are the conversion methods in order to compare these two sensors?