LULC change detection & Prediction with MOLUSCE Tool in QGIS

[@ThompsonSit]

Remember to re enable plug in if the MOLUSCE crashes also why it doesn’t read the TIFF properly

[@ananereatomba9763]

Molusce is not available in QGIS 3.22, do you know about some similar plugin that i can use?

[@nextgis_channel]

Hello! MOLUSCE 4 has been released, now with support for the latest versions of QGIS

[@SijingWong]

Hello，what about the"Number of simulation iterations"? Will the increase of this value lead to more accurate results?😃

[@نذيرصلاحالدينأيوب]

Please, in which step we can specify the required prediction year?

[@gisrssol.6851]

@@نذيرصلاحالدينأيوب it’s in the last step where I am writing down the year

[@none37992]

In the file name?​@@gisrssol.6851

[@none37992]

In the file name?

[@gisrssol.6851]

@@none37992 yes

[@none37992]

Another question please, the resolution of the input images must be the same? And for example, can i use images of 2002 and 2008 predict image on 2023?

[@lucastheodori7370]

Thank you for nice video, i have done it and i want to claculated the area of changed land cover, to to go about

[@gisrssol.6851]

just convert the area of raster into vector and calculator the individual classes area

[@liliana3652]

hi, i m having problem with my data not matching, i have clip and resample the raster DEM and euclidean road distance, bt still not matching. I tried many times but still having different pixel issue, such as one is 3055and another is 3050. will u b able to help? thank you

[@gisrssol.6851]

Hi, you should try some online portals to match your spatial extent and pixels that would be helpful sometimes

[@liliana3652]

thank you for your reply. can you share the web address to align this data?@@gisrssol.6851

[@henokgube3789]

i faced similar problem, how can I fix it

[@akashbehera9700]

kindly provide the older version of QGIS

[@gisrssol.6851]

Okay I will share the link

[@akashbehera9700]

@@gisrssol.6851 thank you so much maam

[@akashbehera9700]

@@gisrssol.6851 maam i am still waiting for the link

[@sharjeelrasib8017]

This tutorial is quite helpful but molusce is not available in the latest version how can we fix this problem?

[@gisrssol.6851]

Use previous version

[@nextgis_channel]

Hello! MOLUSCE 4 has been released, now with support for the latest versions of QGIS

[@aratrika7915]

There's no significant difference between the predicted and previous land use/land cover (LULC) data. What should be the next steps?

[@gisrssol.6851]

@@aratrika7915 which years of data you have utilised? Also do add other layers as well to predict the data future accurately likewise road network

[@Naturistic0786]

This plugin is not available what should I do now.

[@gisrssol.6851]

I have the Google drive link I will share

[@Naturistic0786]

@@gisrssol.6851 please share the qgis old software version link cause I couldn't find it.

[@Naturistic0786]

@@gisrssol.6851 can you share how I can find the older version of Qgis which is compatible with this plugin.

[@gisrssol.6851]

@@Naturistic0786 sure

[@akashbehera9700]

@@gisrssol.6851Can you share it plzz

[@anakalamaheswarareddy7214]

please share the link of qgis version in which molusche plug in is theere

[@gulzamanwilliam3570]

Very nice video...thanks. But I use qgis of the same version that you are using. But unfortunately correlation analysis is not working.. An error occurred..

[@gisrssol.6851]

I have installed my molusce tool by separately through python. That's why it is working properly. Second reason is that may be your units of layers are not same that's why correlation can't be generated

[@gulzamanwilliam3570]

@@gisrssol.6851 thank you so much for your reply. Please can you make video how to install separately through python and how to manage layers..please. My some friends also facing this issue. It will be your great favour. Thanks

[@gisrssol.6851]

@@gulzamanwilliam3570 sure

[@gulzamanwilliam3570]

@@gisrssol.6851 thanks 😊.. I'll wait

[@jobmalanguis3694]

@@gisrssol.6851 hello, are there any updates on the video tutorial? it would be of great help for my undergrad thesis. thank you

[@studymindphd]

What is this yellow background behind. The map? This happens to me as well. Will it affect the area of the class represented by yellow color

[@gisrssol.6851]

You can easily delete it by clip tool

[@mdnahinalam7339]

How do knpw that predicted map is for 2030?

[@gargichandane9670]

I am unable to find the plugin

[@jalanigamer]

Hello how do i fix this problem when I was creating changes map this will happen Traceback (most recent call last): File "C:/Users/63966/.qgis2/python/plugins\molusce\moluscedialog.py", line 430, in changeMapDone self.inputs["changeMap"].save(self.inputs["changeMapName"]) AttributeError: 'NoneType' object has no attribute 'save'

[@ThompsonSit]

3.38.1 Doesnt work for me. Reinstalled 3.38.3 and now works

[@samariagarmagarcia2905]

Hello, can it be used for several years or should it only be two? Also, the information to use are land use layers?

[@meeeerv_8113]

Hi! The plugin isn't available in QGIS 3.28 Firenze 😅 What should I do?

[@gisrssol.6851]

I will upload and share the link so you can download easily

[@akashbehera9700]

@@gisrssol.6851the link is not available in the below link can you provide

[@lordnazareth3165]

Hello. Thank you for the tutorial. I do not see where you define 2030 as the prediction year. I just see how you save a file with the name 2030 before running the prediction. Where can we set the prediction year, please? What is I want to predict for 2025 or 2035 instead of 2030?

[@gisrssol.6851]

whenever we save the file during prediction plugin will automatically pick up the year you can specify your requirements during as like it picks up the initial and final year during start of molusce tool

[@lordnazareth3165]

@@gisrssol.6851 thank you

[@dhruvsaxenadstar]

@@gisrssol.6851 so, if i mention "2040" in file name, will it generate LULC prediction for 2040 ?

[@GK_594]

Please suggest some suitable website for downloading the landuse map of different places from different timeline in India

[@gisrssol.6851]

Landuse maps directly not available but for processing the dataset at different times you can use Google Earth Engine

[@GK_594]

@@gisrssol.6851 thank you

[@naveenkumars5771]

Sir how to install molusce plug in

[@wuenakloe]

The MOLUSCE plugin isn't available in the newer versions of QGIS. How do you work around it in this situation?

[@gisrssol.6851]

use previous version

[@nextgis_channel]

Hello! MOLUSCE 4 has been released, now with support for the latest versions of QGIS

[@shashankdwivedi8796]

mam if i save the file as 2050 prediction.....it will automatically generate 2050 prediction map ??

[@gisrssol.6851]

Yup

[@shunrei7461]

Hi, could you inform us on how did you manage to get the MOLUSCE plug-in please ? thank you!

[@gisrssol.6851]

Sure I am working on that. As I have downloaded this version one year ago but now it's not available online but whenever I got this tool I will upload.

[@geneyan]

@@gisrssol.6851 Hello, do you have an update for this?

[@gisrssol.6851]

@@geneyan I couldn't find it yet.

[@temporary870]

In Transition Potential Modelling Part When I'm Training Model My Current Validation Kappa Is 0 Even It Is Not Moving Also Where In Your Case It Is 0.95171. Can you please enlighten me on what the possible reason is?

[@gisrssol.6851]

I didn't came across such scenario but may be there is some issue with your supervised classification. Can you do it again may be that would be helpful.

[@ujjalsenapati8947]

I am requesting a copy of the QGIS 2.8.9 software and the MOLUSCE plugin.

[@zualteizote6911]

Thank you

[@ulfiistiningdiah381]

Thank you for the good tutorial, but i have a problem when inputting data the result is the geometry is not matched. can you provide a tutorial how to fix the geometry problem between LULC and spatial variables to perform MOLUSCE analysis in QGIS?

[@gisrssol.6851]

sure will do so. Always remember to have similar spatial extent, resolution and geographical coordinates

[@ulfiistiningdiah381]

@@gisrssol.6851 I've previously matched the coordinates and resolution but maybe there are some mistakes

[@ulfiistiningdiah381]

@@gisrssol.6851 Thank you, I'll be waiting for the new video 😊

[@gisrssol.6851]

There are some online platforms also available to match the extent and coordinates. You can try that option!

[@olanikekafayat7057]

Please kindly help us with this molusce Plugin.. Thanks

[@gisrssol.6851]

What sort of help do you need

[@kafayatkolenich8842]

@@gisrssol.6851 I was unable to install MOLUSCE Plugin on Qgis

[@kafayatkolenich8842]

@@gisrssol.6851 Please which Qgis version can I get that MOLUSCE Plugin or can install it. thanks

[@gisrssol.6851]

@@kafayatkolenich8842 2.18.9 version

[@emretunal3590]

Hello. Thank you for the video. QGIS crashes constantly during the Cellular Automata Simulation phase. The change estimate starts at 4% and the application closes. Do you have any idea about this situation?

[@gisrssol.6851]

Please check your internal memory of the computer or a laptop

[@emretunal3590]

@@gisrssol.6851 Thanks for your feedback. Actually my internal memory is working fine. But I will check again.

[@gisrssol.6851]

@@emretunal3590 yes please because sometimes we installed gis softwares our antivirus remains on which also creates trouble while doing analysis

[@novika8708]

I have used all possible combinations of spatial variables like DEM, slope, aspect, distance to roads, urban areas, population etc. but the value of validation kappa is not improving beyond 0.45. Could you suggest how I may improve the validation accuracy in MOLUSCE? all my LULC maps have been prepared with accuracies above 90%.

[@gisrssol.6851]

Sometimes there are some outliers which need to be identified so to increase accuracy . I would recommend to find those outliers by seeing the spectral differences in ENVI software if possible.

[@angelografie9893]

You may also choose to change your hyperparameters when training your ANN model

[@Naturistic0786]

@@gisrssol.6851 how did you install the older version of qgis

[@gisrssol.6851]

@@Naturistic0786 by searching it's version online

[@daenielmojica8242]

did you manage to increase your current kappa validation? Mine also has lower than 0.45

[@dhruvsaxenadstar]

upon adding DEM, Slope, Roads, Urban variable maps, I'm unable to create change map. I'm getting this error. Can you please help ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

[@gisrssol.6851]

Are there coordinates and extent is same ??

[@dhruvsaxenadstar]

@@gisrssol.6851 Yes

[@gisrssol.6851]

@@dhruvsaxenadstar If you are using two variables is it still not working and secondly when you add latest year of LULC in spatial variables is it working or not?

[@dhruvsaxenadstar]

@@gisrssol.6851 Nope. I guess I'll reinstall all software and plugins to try again..

[@badriefendi152]

Thx u for sharing great video! Anyway i need ur help, i want ask some question how to increase validation kappa ? Bcs im stuck in 66% thx u

[@gisrssol.6851]

Use more training sets for each class to increase kappa results

[@badriefendi152]

@@gisrssol.6851 thanks for answering. I am a new student in remote sensing, how to do that?

[@gisrssol.6851]

You should collect more than 50 training samples for one class likewise buildup area so to increase its accuracy and precision while performing supervised classification

[@badriefendi152]

@@gisrssol.6851 thx u i will try it! Have a nice day

[@sarasameh6607]

​ i used LULC maps for 2011 as intial and 2011 as final map and used 2021 as validation map but when try to predicat 2031 get the same map as final map(2011) while i use DEM and distance from road and slope as independent variables? can youy help me to fix yhis problem?

[@gisrssol.6851]

Use 2021 as final map map and use other spatial variables for validation

[@نذيرصلاحالدينأيوب]

when i add slope map as a spatial variable to predict LULC by Weight of evidence i face a problem that the Rand min and Range max for the slope raster appers as (nan) how can i fix this problem, please?

[@anakalamaheswarareddy7214]

please tell me the solution for probelm am facing (after adding classified images and spatial variables am getting answer as geometrics are not matched input data and spatial variable ,please tell me how to solve

[@gisrssol.6851]

Sometimes even if we give same coordinates to raster images still there is an issue of a spatial extent. I would recommend you should convert your shapefiles and extent with similar properties by converting it from some online websites. That would be helpful more and the geometrics issues will be resolved.

[@akashbehera9700]

This plugin is not available what should I do now

[@nextgis_channel]

Hello! MOLUSCE 4 has been released, now with support for the latest versions of QGIS