Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Karthikeyan L

Pages: [1] 2 3 ... 5
Data / Free GIS Data
« on: March 11, 2019, 11:38:41 PM »
The following contains information on variable wise as well as country wise free GIS datasets. A very good summary!

Professor PP Mujumdar, Department of Civil Engineering, Chairman, Interdisciplinary Centre for Water Research, Indian Institute of Science, Bangalore, has been elected as the fellow of Indian National Science Academy (INSA). Prof. Mujumdar is the first hydrologist to get this recognition.

Here is the full list of elected members:

Prof. Mujumdar's webpage can be accessed through this link:

A work carried out by Sonali Pattanayak titled "Linkage between global sea surface temperature and hydroclimatology of a major river basin of India before and after 1980" got published recently in the Environmental Research Letters.


Recently, I have published a couple of papers related to satellite microwave soil moisture research in the Advances in Water Resources Journal.

The process of retrieving soil moisture from satellite microwave sensors depends on the type of sensor i.e., active and passive microwave sensors. Over the past four decades, the microwave community has progressed in terms of improving the sensor design and retrieval algorithm so as to achieve accurate global scale soil moisture observations.

The first paper titled "Four decades of microwave satellite soil moisture observations: Part 1. A review of retrieval algorithms" gives a comprehensive overview of the developments that took place in the retrieval algorithms over the past four decades. The paper discusses the algorithmic developments of both active as well as passive sensors. We have also, for the first time, summarized the literature in the form of figures, one each for active and passive microwave soil moisture research (PFA). This review also discusses the latest developments in components of the algorithms, and also discusses the challenges that need attention in future. I can say that this paper serves as a starting point for someone who wants to venture into the field of microwave soil moisture research.

The second paper titled "Four decades of microwave satellite soil moisture observations: Part 2. Product validation and inter-satellite comparisons" is focused on assessing the accuracy of soil moisture products, which were developed over the past forty years. We have considered eight passive (SMMR, SSM/I, TMI, AMSR-E, WindSAT, AMSR2, SMOS, and SMAP), two active (ERS-Scatterometer, and MetOp-ASCAT), and one active-passive combined (ESA-CCI combined product) soil moisture products for the analysis. The Contiguous United States (CONUS) is considered as a case study. The validation is carried out using the data pertaining to 1058 stations over the CONUS, along with model soil moisture simulations obtained from the VIC land surface model. We analyzed these products in terms of daily coverage (a figure in this context is atteched with this post), temporal performance, and spatial performance. We also carried out inter-satellite comparisons to study the roles of sensor design and algorithms on the retrieval accuracy. The Part 1 serves as a prelude to this paper. Through these papers, one can get an idea about how well the satellites along with algorithms have progressed over the four decades and significantly improved the accuracy of soil moisture retrievals.

Part 1 can be downloaded for free from here: (until November 4, 2017)
Part 2 can be downloaded for free from here: (until November 17, 2017)

Here are the permanent links:
Part 1:
Part 2:

You can write an email to me at for full-text requests or any other query related to papers.

The work carried out by Ila Chawla along with Prof. P.P. Mujumdar titled "Partitioning Uncertainty in Streamflow Projections under Nonstationary Model Conditions" was recently published in the Advances in Water Resources Journal.

In this work, the authors, in a novel attempt, have addressed the possibility of nonstationary hydrological model (here they have used the VIC land surface model coupled with a routing model). They found that the hydrological model parameters, which are influenced by climate variables, vary over time, and thus may not be assumed to be stationary. Further, the work also involves attribution the total uncertainty in the streamflow projections to multiple effects such as, model parameters, GCM simulations, emission scenarios, land use scenarios, the assumption of hydrological model stationarity, along with internal variability of the model. The Upper Ganga Basin (UGB) is considered as a case study for this analysis.

Further details on the work can be found here:

Authors are glad to share their article with interested readers.

Correspondence to:

Programming / Re: Exporting High Resolution figures from 'R'
« on: July 06, 2017, 05:28:03 PM »
And I forgot about 'Cairo' error. With the current version of Cairo, you cannot create tiff file. Type Cairo.capabilities() to see what all formats are supported. To make this work, you need to edit the source code of Cairo library using the modified code put up by the Cairo developers on this github page If you wan to avoid all of it, then go with default function tiff. Apart from smoother plots, the only other advantage with Cairo is the option of enabling background transparency.

Programming / Re: Exporting High Resolution figures from 'R'
« on: July 06, 2017, 04:59:58 PM »
Give the width and height units in inches. It will be easier for you to understand and manipulate the values. Plot error regarding margins could be due to insufficient space in the RStudio window. I know its silly, but you can increase the space for the plot by dragging with mouse. Also, I can see that you are not using command, you should include that too. Forget the pointsize for now, increase the resolution, specify width and height in inches, and let us know what you have got!

Code: [Select]
tiff('test.tiff', width=20, height=20, res=300)----> The default units are pixels, so you are giving the size as 20 x 20 pixels, which is definitely too small for the figure to plot, isn't it? So, my advice is to change the units to inches and then keep playing with widths and heights

If you are still unable to plot, please provide data here, we can try to help you with that!

Programming / Re: Exporting High Resolution figures from 'R'
« on: July 06, 2017, 12:21:18 AM »
We can create high resolution plots in two ways: 1) using the default functions; 2) using the 'cairo' package. Say you want to create a tiff file. Both methods have the following syntax,

Code: [Select]
tiff (filename='filename.tiff',  width=5, height=4,units="in",pointsize=12,res=72)
$$your plot command$$ ()
In case of cairo, load the library first, instead of tiff we use:
Code: [Select]
cairo(filename='filename.tiff', type="tiff", width=5, height=4,units="in",pointsize=12,res=72)

Models / Re: I need help to run Skehekin sample with VIC 5.0
« on: June 24, 2017, 05:36:34 PM »
Also, please look into the other threads posted in this forum regarding the VIC model. To avoid redundancy, start new thread only when you are not able to find answer to your question in our older threads.


Models / Re: I need help to run Skehekin sample with VIC 5.0
« on: June 24, 2017, 05:34:05 PM »
You will have to compile the model first. You can find detailed explanations regarding setting up the VIC model in the following links:

Post your question/information / Re: Indexed image in Matlab
« on: May 11, 2017, 11:48:36 AM »
The example that I have given towards the end is for the indexed image. 'corn.tif' is an indexed image.

Post your question/information / Re: Indexed image in Matlab
« on: May 10, 2017, 10:23:00 AM »
An indexed image is useful to represent the image which has limited number of colors in it so as to reduce the space consumption (of the image). As the name suggests, an indexed image is a matrix of indexes with each number pointing to the color in the colormap matrix.

Say you have an image with only three colors red, green and blue and no other color, its indexed image X (assuming it to be a 4x4 matrix image) could look something like this:
Code: [Select]
X=[3,2,2,1;1,3,1,2;1,2,1,2];And the corresponding colormap matrix will be:
Code: [Select]
map=[1,0,0;0,1,0;0,0,1];Here map will be 3x3 matrix with rows indicating number of colors (which is three in this example) and columns indicating R,G,B (the numbers will be varying between [0,1]).
For example, 3 in X simply points to 3rd row in map matrix and takes its corresponding color (which is blue).

Once you obtain index and colormap from an image, you can use them to get an RGB image by using ind2rgb
 ( command. This command simply maps indexes to the colormap matrix and creates a 3D matrix where the third dimension contains the RGB colors. You can run following lines in matlab for better understanding:

Code: [Select]
[X,map] = imread('corn.tif');
RGB = ind2rgb(X,map);

« on: April 10, 2017, 10:26:29 AM »
You can do a simple google search and find out for yourself the answer to your question. For your assistance, here is the link of google search results:
Let us know what you found from the search and then we can discuss further.

If you are dealing with huge datasets that are saved in .mat format, loading them to workspace eats lot of RAM. So, matlab has an option to read part of such files without loading them to memory. Have a look at this link:

Interesting information / A lecture by Prof. Fawwaz Ulaby
« on: January 28, 2017, 11:40:36 PM »
In this lecture Prof. Fawwaz T Ulaby, a pioneer in Radar Remote Sensing talks in a fascinating way about how Radar technology was developed in 1970s. You can see the first half of the video.

Pages: [1] 2 3 ... 5