C6 data in ftp://ladsweb.nascom.nasa.gov/allData/6/

Note that C6 is different than C5.1 in the way that AOD is processed, at least in the monthly L3 data I use. The DB and DT results are combined into a single AOD value called "AOD_550_Dark_Target_Deep_Blue_Combined_Mean_Mean".

Dear Ram,

I think you can get some information from this link

http://gdata1.sci.gsfc.nasa.gov/daac-bin/G3/gui.cgi?instance_id=MODIS_DAILY_L3

This is the Giovanni NASA website from where you can choose your AOD data as per your requirement.

Here you can also visualize the data sets as well as can download datasets in HDF or ASCII format.

HOPE IT WILL HELP YOU....................

Thank You Mr.Mishra, I will follow your suggestions and I come back to you. . if i have any queries on that..

Dear Ram,

The Giovanni website is the way to go. It is fairly straight forward. I usually download the data in NetCDF format which is a self-describing format that can be read by many analysis and visualisation packages (e.g., IDL, Matlab, GrADS, etc.)

If you work over bright land surfaces such as deserts then the MODIS Deep Blue AOD is your preferred option. This algorithm was developed to work well over deserts.

Hope this helps,

Sebastian

Dear Sebastian,

Thank you for your suggestion, It gave a broad view of its application

Well the prime problem arises after downloading how to plot them and extract the desired results. What can be done in that case?

To analyse MODIS data, You can use some software to read NetCDF files (e.g. ENVI, MATLAB). Apart from these softwares You can also use your own programmes in FORTRAN, IDL etc. To vizualization of your desired type (e.g. contour plot on map, time series averged over area lat/lon, or any statistical anayses), You can use any abovemetioned softwares or programming language. You can also get the control file to read the pescribed format (NetCDF/CDF) online according to your choosen programming langauges (IDL, FORTRAN etc.).

Hope It will help You..............good luck

Thank You Dr. Mishra as per the suggestions given by You and Dr, Sebastian I am planning to use MATLAB AND FERRET for my analysis.

But i am having doubts on algorithms for retrievals I am seraching some reprints for suitable algorithms used by researchers .

Dear All, please find a new Simplified High Resolution (500m) MODIS Aerosol Retrieval Algorithm (SARA):

http://www.sciencedirect.com/science/article/pii/S0034425713001417

Dear Parth

NetCDF file format has its protocol

You can find and study it.

If you need the program urgently, send an e-mail to me at "[email protected]", then I will send you a simple MATLAB program to read and visual deepblue AOD data in .nc format.

Have a good time

Moradhaseli

Here I have attached a sample output image.

You can also use the MISR data at 555 nm, which in some cases is of better quality then MODIS. This is also normally available on Giovanni, but the site is currently unavailable due to the US government shutdown.

Chris is right we have similar observation over desert MISR 555 works well. I am not sure if MODIS V6 is available we were planning to try that for AOD reterivals.

MODIS AOD v6 is not available yet, at least on Giovanni...

Back to the initial question, don;t forget that there are a lot of missing pixels in all aerosol observations from satellites. For applications where you need complete temporal records, that is a huge problem.

Chris do you see any problem if these missing data are generated by extrapolation.

Hello everybody, we started working with MODIS Level 3 v5.1 some years ago but discovered that it has a lot of missing data and overstimated AOD in many locations by a factor of 2. After that, we switch to MACC and solved the problem of missing data temporaly and spatially but as it uses MODIS data as input the overestimation is maintained. Now, we are having better results with MISR. Its estimations of AOD are more realistic and the number of missing data is minimum, almost for the places where we have worked with it (China, Uzbequistan, India, North of Africa).

Saif, extrapolation is usually not a good solution, if at all possible. One important factor is the geographic area of concern. Over arid areas, MODIS considerably overestimates AOD. For instance, over the southwest USA, it overestimates by 4 to 10 times on average! MISR is only half-way better there.

MACC has no missing pixels (because modeled), but starts only in 2003 and has cruder spatial resolution. Since the model is constrained by MODIS observations, it also suffers wherever MODIS is inconsistent, as also mentioned by Luis.

For the best results in solar irradiance modeling, my company sells monthly-effective AOD data at medium (0.5x0.5°) to high spatial resolution (0.1x0.1°), with no missing pixels (since 2000), and with the least possible bias over arid areas. See sample below for North America in April 2004.

Yes we have experienced similar situation in Kuwait and Gulf Countries. MODIS overestimates 2 - 6 times.

Chris, your database looks awesome. It is of great value having unbiased data with that spatial resolution and from 2000.

Chris I should say am impressed, can I have a look over Persian Gulf, if you have a sample and how much your company charges for a monthly effective AOD at 0.1 x 0.1 resolution

Saif, please contact me directly through LinkedIn, for instance

Can I please have your email Chris mine is [email protected]

I have tried to send you mail but failed . Might be your given email id has some missing. Could u kindly give your right mail? My mail id : [email protected]. Actually can i get Deep Blue AOD read and visual mathlab program? sorry for the past some spelling mistake.

Arfan, With whom do you want to be connected?

Saif, I was recently at KISR under an extremely tight schedule to give a workshop on solar resource assessment for delegates from MENA. The aerosol impact on the solar resource was obviously an important topic of discussion. I will certainly be visiting KISR on a regular basis, so we'll have the chance to meet in person.

Thank you very much to Chris Gueymard. Actually I have tried to say to Ruhollah Moradhaseli.  

Salam Yasmin -

I have used MYD04_3K for Beijing and found that it has low retrieval quality/accuracy than MYD04 10K. It also have missing pixels due to pixels selection criteria/thresholds of Dark-Target (DT) algorithm, 

Dear Dr. Bilal,

In which sense the MYD04_3K has low retrieval quality? I have been using it since its availability and found very nice correlations in indo-gangetic plains with ground based AOD from microtops which was much better than MYD04_10K or MOD04_10K. What are the other things which you found of lower quality?

Thanks

Dear Manish, 

Did you evaluate the MYD04_3K and MYD04_10K AOD observations using Expected Error [+/- (0.05 + 0.15AOD)]? Yes, MYD04_3K C6 almost has the same correlation as MYD06_10K C6 with the AERONET AOD measurements, but AOD observations from MYD04_3K are more overestimated than MYD04_10K AOD observations, as most of the observations fall above the upper envelope of EE. Overall, both MYD04_3K and MYD04_10K have only

Dear Dr. Bilal,

Yes you are right accuracy is more important than correlation. I did not evaluate the two products. My observations were based on ground based AOD and correlations between MYD04_3k with surface monitored particulates. I will try to figure out it again. So what will you suggest to those researchers who are consumer of these products for different aspects of aerosols but not interested in detailed algorithms. Suppose one has to analyze the relation between PM2.5 and satellite AOD, which product would be better? 3km or 10km? Because surface monitoring using dust samplers represent a limited area and AOD associated with larger area i.e. 10x10km would not help to study the relations precisely. Please suggest.

Regards,

Manish ([email protected])

Dear Manish,

I will recommend to evaluate the retrieval accuracy using EE before using for any further research application. For example, to develop a regression between AOD and P2.5. According to my experience over Beijing and studies for MYD04_3K, the retrieval accuracy of MYD04_3K is low as compared to MYD04_10K, but it may vary region to region. It is better to map PM2.5 at higher spatial resolution but first need to evaluate the higher resolution product for particular region. By the way, MYD04_10 K DB C6 has almost 3 (7) times better accuracy than MYD04_10K DT C6 (MYD04_3K DT C6) over Beijing, so you can also consider DB C6 if you want to use AOD at 10 km resolution.

You can also check my AOD algorithm at 500 m spatial resolution: 

http://www.sciencedirect.com/science/article/pii/S0034425713001417

http://www.sciencedirect.com/science/article/pii/S0034425714002636

Regards,

Bilal

Thanks Dr. Bilal for your suggestions. I will communicate with  you further in case of any other query.

Regards,

Manish

I said if someone want to use AOD at 10 km spatial resolution then I would recommend to use DB AOD C6 rather than DT AOD C6. Overestimation or underestimation depends on the land cover types and aerosol types. As MYD04_3K is not thoroughly validated  over different regions so I will recommend you try to validate MYD04_3k first and then decide which aerosol product is useful for your region. If you don't want to validate MYD04_3K and you think that AOD at 10 km resolution is suitable for your research then you can go for DB C6 AOD at 10 km.

Regards

Bilal

Dear Sir

This is Chandrasekar doing research in Air quality model. In part of my research I am using MODIS AOD data. MOD04_L2.A2015044.0200.006.2015044170006.hdf. Please tell me the time in the file is UTC or local time of the over pass.

Thanks

Hi,

In this file "MOD04_L2.A2015044.0200.006.2015044170006", UTC time is 0200 (2am).

Regards,

Bilal

Dear sir

where can i download DB AOD C6 data

thanks

C6 data in ftp://ladsweb.nascom.nasa.gov/allData/6/

Note that C6 is different than C5.1 in the way that AOD is processed, at least in the monthly L3 data I use. The DB and DT results are combined into a single AOD value called "AOD_550_Dark_Target_Deep_Blue_Combined_Mean_Mean".

To gain more information about this product,you can refer to the following publication:

Article Investigation and validation of MODIS SST in the northern Persian Gulf

Dear All,

I need AOD monthly data for ocean. From the literature it seems that "Effective Average Ocean" is the best SDS for ocean AOD. But how to download monthly AOD from LAADS website. NASA Giovanni website does have this parameter. Looking for suggestions. 

AOD data at different temporal resolution is available at earthdatasearch (https://search.earthdata.nasa.gov/search?m=-24.1875!-125.71875!0!1!1!0%2C2%2C1).

follow the link

https://ladsweb.nascom.nasa.gov/search/

Dear all

I need to determine the quality accuracy of the L2 AOD 550 nm data (Deep Blue and Dark Target) before processing data. What quality accuracy I have to use for Deep Blue and Dark Target data?

You can dispose the MODIS products using IDL, NCL, python and Matlab. In addition, you can preview the data using HDFview or HDEexpler. Wish this help you!

Thanks for your help. I have been looking for information and I have been able to work with the MODIS data.

Thanks for your help

You can refer Ladds web page information. And to extract AOD after you download the data from that web, MATLAB would be recommended to first change it to NETCDF data format in order to further analyze that data using ncwriter and modis reader Matlab scripts.

Is it possible to transfer the MODIS C6.1 data into NetCDF file, though the significant amount of missing data in MOIDS?