search

eurodatacube / eodash

Software behind the RACE dashboard by ESA and the European Commission (https://race.esa.int), the Green Transition Information Factory - GTIF (https://gtif.esa.int), as well as the Earth Observing Dashboard by NASA, ESA, and JAXA (https://eodashboard.org)
https://race.esa.int
MIT License
93 stars 42 forks source link

Integration of data from CCI lakes #2202

Closed santilland closed 10 months ago

santilland commented 1 year ago

Source: https://climate.esa.int/en/projects/lakes/data/ CCI has shapefile for global lakes, and csv with names. NetCDF data - example notebook to retrieve data Need for conversion to COG?

Parameters of interest:

Maybe of interest: https://collections.eurodatacube.com/clms-lake-water-quality/

santilland commented 1 year ago

LAKES:

Tonlè Sap Lake (in Cambodja) Varnen Lake (in central EU) Balaton Lake (in central Hungary) Biwa Lake (in Japa) Colorado River Aral Sea

aapopescu commented 1 year ago

For the lakes in the list, we need to add this data:

NASA :

ESA:

JAXA:

santilland commented 1 year ago

For EOX - extract "subaoi" from lakes shapefile and create indicator. For some lakes s2 and other sources might be used, whatever source we use we would need a list of dates where source is available over the lake extent

aapopescu commented 1 year ago

@gunbra32, we'd like to use the the following CLMS dataset in RACE and we'd need your support to access this via EDC. https://collections.eurodatacube.com/clms-lake-water-quality/

Cc: @AlessandroScremin

santilland commented 1 year ago

Hello @j08lue, we are working on the integration of data for the trilateral eodashboard preparing for a lake story that is expected also to be used as possible example for foss4g workshop. In order to have data also from NASA could you help us locate data from either Landsat, HLS, PlanetScope that covers the lakes listed further up?

j08lue commented 1 year ago

could you help us locate data from either Landsat, HLS, PlanetScope that covers the lakes listed further up?

I will look into that, asking whether there is a preference on the NASA side for which data to provide.

Questions:

  1. Do we have AOIs for these lakes somewhere already?
  2. What time range and temporal resolution would be preferred?
  3. Any other criteria for the scenes to present - low cloud cover or something?
AlessandroScremin commented 1 year ago

@j08lue here are the AOIs centroids points..

Tonle Sap (Boeng Tonle Chhma - 12.793990, 104.154704 (Cambodia) Vanern - 58.868459, 13.344541 (Sweden) Balaton - 46.838203, 17.712136 (Hungary) Biwa - 35.301566, 136.133431 (Japan)

j08lue commented 1 year ago

I will try to find the story text - it is probably already written up - so I understand what kind of data would add most value to your story.

In particular, what would you want to show with Landsat or HLS surface reflectance? The general landscape around the lakes? Or an evolution of features seen on the lake (ice, algal blooms or so) over specific periods?

For example, here is a nice cloud-free Landsat-9 scene of the Vännern lake on 8th of May 2023:

image

(And a Microsoft Planetary Computer tiler URL: https://planetarycomputer.microsoft.com/api/data/v1/item/tiles/WebMercatorQuad/{z}/{x}/{y}@2x?collection=landsat-c2-l2&item=LC09_L2SP_195019_20230508_02_T1&assets=red&assets=green&assets=blue&color_formula=gamma+RGB+2.7%2C+saturation+1.5%2C+sigmoidal+RGB+15+0.55&nodata=0&format=png)

If you would prefer to get PlanetScope scenes, we can probably also get those. In the COVID Dashboard API, we had a proxy for selected PSScenes. That is not part of the VEDA API, but maybe we can find a different mechanism to provide them, if needed.

AlessandroScremin commented 1 year ago

@j08lue nice to see it with landsat. What we are going to provide are 3 indicaotrs related to the lake: water surface temperature timeseries and maps water lake extent timeseries water level timeseries

I don't think with landsat you are able to retrieve all those information. Maybe surface temperature. Lake extent I see too diffult with the spatial resolution of Landsat.

At leat this was the idea to be onfrmed by @aapopescu

AlessandroScremin commented 1 year ago

@aapopescu @santilland @lubojr

these are the information list I can retrieve from ESA Lake project:

here an extract for water extent ovemposed with water level

image

aapopescu commented 1 year ago

@gunbra32, we'd like to use the the following CLMS dataset in RACE and we'd need your support to access this via EDC. https://collections.eurodatacube.com/clms-lake-water-quality/

Cc: @AlessandroScremin

Follow-up on this, maybe @AliceBalfanz can help. Thanks!

pont-us commented 1 year ago

The LWQ 300m dataset is currently available in an S3 bucket, but only for the time range 2018-01-1 to 2020-12-21:

from xcube.core.store import new_data_store
data_store = new_data_store('s3', root='cop-services')
lwq = data_store.open_data('LWQ-NRT-300m.zarr')
aapopescu commented 1 year ago

I will try to find the story text - it is probably already written up - so I understand what kind of data would add most value to your story.

Hi, @j08lue , the story is drafted at this link and you should have access https://docs.google.com/document/d/1Nc0bZhd6Myz6kTZvXPSgkJamYN4iHjm7D-lpCOtCx8k/edit

In particular, what would you want to show with Landsat or HLS surface reflectance? The general landscape around the lakes? Or an evolution of features seen on the lake (ice, algal blooms or so) over specific periods?

We are looking to show the variation in the water level and water quality, so what we need to visualise is:

For example, here is a nice cloud-free Landsat-9 scene of the Vännern lake on 8th of May 2023:

image

(And a Microsoft Planetary Computer tiler URL: https://planetarycomputer.microsoft.com/api/data/v1/item/tiles/WebMercatorQuad/{z}/{x}/{y}@2x?collection=landsat-c2-l2&item=LC09_L2SP_195019_20230508_02_T1&assets=red&assets=green&assets=blue&color_formula=gamma+RGB+2.7%2C+saturation+1.5%2C+sigmoidal+RGB+15+0.55&nodata=0&format=png)

If you would prefer to get PlanetScope scenes, we can probably also get those. In the COVID Dashboard API, we had a proxy for selected PSScenes. That is not part of the VEDA API, but maybe we can find a different mechanism to provide them, if needed.

Higher resolution imagery is not essential but could be useful. Would it be feasible within the next 2 weeks?

santilland commented 1 year ago

@leahsturm to provide list of s2 acquisition dates (taken from https://apps.sentinel-hub.com/eo-browser/) looking for good visibility (as little cloud cover as possible) and where changes in lake extent can be seen --> EOX to add s2 indicator with dates for lakes with provided dates

santilland commented 1 year ago

Data to be explored per lake: Aral sea - extension (S1 or S2) Lake Biwa - extension (S1 or S2) , temperature (S3 if res. enough) and quality (S2) Colorado River (and Lake Powel) - extension Tonlé Sap - level Lake Varnen - water quality (S2 or S3 if resolution is enough) Lake Balaton - ice cover comparison between different winters (S1)

AlessandroScremin commented 1 year ago

The LWQ 300m dataset is currently available in an S3 bucket, but only for the time range 2018-01-1 to 2020-12-21:

from xcube.core.store import new_data_store
data_store = new_data_store('s3', root='cop-services')
lwq = data_store.open_data('LWQ-NRT-300m.zarr')

Dear @pont-us @gobaRules @gmilcinski We would need to have these datasets avaialbe on my account in SH for the Race dashbaord. You (brockman/Synergize) already do that for me with the BICEP dataset, would be possible to do that also in this case?

gmilcinski commented 1 year ago

@Dorothyrono - can someone in your team take care for ingestion of this collection in the Sentinel Hub? And/or provide to Alessandro the script so that he can do this himself?

gunbra32 commented 1 year ago

@AlessandroScremin @aapopescu @gmilcinski If Sentinel Hub is used, we are obviously out here. Our approach would have been to serve the cube (or just the AOIs) through xcube server (the cube is actually already in the viewer but we would have to optimise it for performance).

aapopescu commented 1 year ago

@AlessandroScremin @aapopescu @gmilcinski If Sentinel Hub is used, we are obviously out here. Our approach would have been to serve the cube (or just the AOIs) through xcube server (the cube is actually already in the viewer but we would have to optimise it for performance).

@santilland , @gunbra32 let's do this exercise. I'd like to demonstrate visualising on RACE data served by the xcube viewer. Concurrently, let's bring this data in SH. Would be great if @AlessandroScremin could have access to the required script. Thanks, @Dorothyrono

gunbra32 commented 1 year ago

Good! This would also be a great use case or the xcube viewer + server activity.

As written above, we will let you know once we optimised the cube. The server provides the following APIs: https://edc-api.brockmann-consult.de/api/openapi.html I assume th WMTS will be used here?

leahsturm commented 1 year ago

@leahsturm to provide list of s2 acquisition dates (taken from https://apps.sentinel-hub.com/eo-browser/) looking for good visibility (as little cloud cover as possible) and where changes in lake extent can be seen --> EOX to add s2 indicator with dates for lakes with provided dates

Here is the list with the Sentinel2 acquisition dates. I noted down the most possible number of cloud free images for each lake to ensure an extensive documentation of the change processes.
Sentinel2_lakes.xlsx

Dorothyrono commented 1 year ago

@AlessandroScremin @aapopescu @gmilcinski If Sentinel Hub is used, we are obviously out here. Our approach would have been to serve the cube (or just the AOIs) through xcube server (the cube is actually already in the viewer but we would have to optimise it for performance).

@santilland , @gunbra32 let's do this exercise. I'd like to demonstrate visualising on RACE data served by the xcube viewer. Concurrently, let's bring this data in SH. Would be great if @AlessandroScremin could have access to the required script. Thanks, @Dorothyrono

@AlessandroScremin The data is ingested in your SH account

collection: Lake Water Quality id: ef59d045-a0cd-4ce0-a25d-5da91e70e965

(Please ignore the ingestion errors that appears in the collection- it does not affect the collection in any way.)

aapopescu commented 1 year ago

@leahsturm to provide list of s2 acquisition dates (taken from https://apps.sentinel-hub.com/eo-browser/) looking for good visibility (as little cloud cover as possible) and where changes in lake extent can be seen --> EOX to add s2 indicator with dates for lakes with provided dates

Here is the list with the Sentinel2 acquisition dates. I noted down the most possible number of cloud free images for each lake to ensure an extensive documentation of the change processes. Sentinel2_lakes.xlsx

Thanks, @leahsturm, this is great! @lubojr (Cc: @santilland ) we should integrate these S-2 images over the lakes in the list. Thanks!

santilland commented 1 year ago

Thanks, @leahsturm created new issue #2251 to keep track of integration of this separately

santilland commented 1 year ago

Good! This would also be a great use case or the xcube viewer + server activity.

As written above, we will let you know once we optimised the cube. The server provides the following APIs: https://edc-api.brockmann-consult.de/api/openapi.html I assume th WMTS will be used here?

@gunbra32 i was not aware of a viewing endpoint for xcube data, with WMTS we should be able to directly integrate it then, not sure about what authentication would be needed for the endpoint though.

The link you provide seems to be configured for a localhost server, so it is not possible to try out things, not sure if this is intended.

In any case let us know once you have optimized the data and which WMTS endpoint we can use to visualize it, thanks!

santilland commented 1 year ago

@AlessandroScremin The data is ingested in your SH account

collection: Lake Water Quality id: ef59d045-a0cd-4ce0-a25d-5da91e70e965

great! thank you @Dorothyrono . @AlessandroScremin could you create a visualization layer based on the data in SH and give us the layer identifier?

AlessandroScremin commented 1 year ago

@AlessandroScremin The data is ingested in your SH account

collection: Lake Water Quality id: ef59d045-a0cd-4ce0-a25d-5da91e70e965

(Please ignore the ingestion errors that appears in the collection- it does not affect the collection in any way.)

Dear @Dorothyrono thanks a lot for the support

I was looking at the visualisation layer as well and at the two layers: Turbidity mean Water qualitu true color.

what is not much clear to me in the true color is if it is binary mask ot what else? this is the result from the layer for true color: this image over north italy central europe

even if the dataMask band has been set there is a lot of white as no data or fill values ...

image

AlessandroScremin commented 1 year ago

great! thank you @Dorothyrono . @AlessandroScremin could you create a visualization layer based on the data in SH and give us the layer identifier?

@santilland these are the layers as provided if any updates will come I will update those layers evelscripts

Lake_Wate_Quality_TrueColor | ID: LAKE_WATE_QUALITY_TRUECOLOR

Lake_water_quality_Turbidity_mean | ID: LAKE_WATER_QUALITY_TURBIDITY_MEAN | 

AlessandroScremin commented 1 year ago

@santilland @lubojr @aapopescu

Dears I created the tables in the GEODB with timeseries (2010-2020) for: Vanern: water lake extent (from CCI lake initiative datasets) Vanern: water lake level (from CCI lake initiative datasets)

the two tables are here: database="eodash" table_name = "Lake_water_extent_tri_timeseries"

database="eodash" table_name = "Lake_water_level_tri_timeseries"

as discussed there isn't a match with any maps, these are only timeseries to be extracted with the chart.

let me know if else is needed or to be clarified. Anyway I'm still working for the water surface temperatures (timeseries)

Alessandro

AlessandroScremin commented 1 year ago

@aapopescu @santilland cc: @dmoglioni

Dear Anca and Daniel with Diego we wrote a JN that work on a standard base image and do the following: -iniect credentials -open a session -explore the SentinelHub Ctalog structure and collections -Filter all the tiles avaialbe over an AOI and with a metadata cloud treshold -print the previews -Compute the statistics to filter Cloud coverage over AOI and acquistion of tiles covering full AOI -print the previews of the "Filtered images"

here in attachment the JN.

follow some final results

First image All images list second image filtered dates based on ccoverage and AOI coverage third iamge print of the previews for the filtered dates

image

image

image

OpenJN_Retrieve_images_over_lakes (1).zip

AlessandroScremin commented 1 year ago

@aapopescu @santilland @dmoglioni

Unfortunately I won't be able to join the weekly meeting tomorrow so I will try to summaryze here the tasks done on our side: DONE

TO DO

lubojr commented 1 year ago

@AlessandroScremin Which is the collection id for the layer LAKES_SURFACE_WATER_TEMPERATURE (so we can extract the list of dates automatically)

@lubojr 97cd622e-40bf-44d6-838b-7ea57a4cc3f5

lubojr commented 1 year ago

@Dorothyrono We tried to extract the list of dates for the client for the given collection using the usual WFS call that we use for other indicators in eodash: https://services.sentinel-hub.com/ogc/wfs/sh-id?REQUEST=GetFeature&srsName=EPSG:4326&TIME=1900-01-01/3000-02-01&outputformat=application/json&TYPENAMES=DSS10-ef59d045-a0cd-4ce0-a25d-5da91e70e965&BBOX=-180,90,180,-90 but it is returning an empty array of features:

{
    "type": "FeatureCollection",
    "features": []
}

Has the functionality been somehow moved or not made available for this one? Is there any other way how we could extract the list of individual dates for configuring our WMS TIME array?

AlessandroScremin commented 1 year ago

@AlessandroScremin Which is the collection id for the layer LAKES_SURFACE_WATER_TEMPERATURE (so we can extract the list of dates automatically)

@lubojr

the id is: 97cd622e-40bf-44d6-838b-7ea57a4cc3f5

lubojr commented 1 year ago

@AlessandroScremin @santilland @leahsturm fyi @aapopescu There has been some progress on integrating the available data to the eodashboard. https://eodash-trilateral-testing.eox.at/cci-lakes-integration/

All new indicators were added as 'oceans' theme for now.

New POIs:

Sentinel 2 L2A over 6 lakes with preselected dates

Looks great! Still missing some markdown story but that is it I think?

Surface Temperature Maps

  • Maps of water surface temperature for:

    • Vanern, Tonle, Biwa and Balaton (visualisation layer--> Lakes_surface_water_temperature | ID: LAKES_SURFACE_WATER_TEMPERATURE

Lake Water Quality

collection: Lake Water Quality id: ef59d045-a0cd-4ce0-a25d-5da91e70e965

Nothing integrated yet as we need a list of dates

AlessandroScremin commented 1 year ago

Sentinel 2 L2A over 6 lakes with preselected dates

Looks great! Still missing some markdown story but that is it I think?

Surface Temperature Maps

  • Maps of water surface temperature for:

    • Vanern, Tonle, Biwa and Balaton (visualisation layer--> Lakes_surface_water_temperature | ID: LAKES_SURFACE_WATER_TEMPERATURE
  • [ ] In the provided collection, there are no scenes over Tonle (bbox query via wfs does not find anything over the tonle bbox or nearby), just the 3 other lakes. Error in ingestion or intention?
  • I have checked several times for individual lakes and the coverage is sometimes very very spotty, but that is a data problem.
  • [ ] Which legend should I reference ? (Which evalscript was used and what are the units)?
  • [ ] Markdown story missing

Lake Water Quality

collection: Lake Water Quality id: ef59d045-a0cd-4ce0-a25d-5da91e70e965

  • [ ] WFS query on this collection does not find any scenes. @AlessandroScremin is this the newest collection for the layers LAKE_WATE_QUALITY_TRUECOLOR and LAKE_WATER_QUALITY_TURBIDITY_MEAN ?

Nothing integrated yet as we need a list of dates

@lubojr Tonle: my fault. No data ignested yet. Will do tomorrow.

Which legend should I reference ? (Which evalscript was used and what are the units)?

Unit are Kelvin. Legend color same for 2m Temperature. The eval script here
'''''//VERSION=3 function setup() { return { input: [{ bands: ["waterTemperature", "dataMask"], // this sets which bands to use }], output: { // this defines the output image type bands: 4, sampleType: "UINT8" } }; }

function evaluatePixel(sample) {

var arr = colorBlend(sample.waterTemperature, [270, 275, 280, 285, 290, 300, 350], [[44,89,150,],[97,154,195],[188,214,230],[246,246,246],[242,200,176],[204,110,90],[146,36,45] ]); if (sample.dataMask==1) arr.push(255); else arr.push(0); return arr; } ''''

WFS query on this collection does not find any scenes. @AlessandroScremin is this the newest collection for the layers LAKE_WATE_QUALITY_TRUECOLOR and LAKE_WATER_QUALITY_TURBIDITY_MEAN ?

yes new collection.

AlessandroScremin commented 1 year ago

@lubojr

Lakes surface temperature dates in the list attached. (added also Tonle even if not yet ingested)

Balaton_dates.txt Biwa_dates.txt Vanern_dates.txt

Dorothyrono commented 1 year ago

@lubojr

https://services.sentinel-hub.com/ogc/wfs/sh-id?REQUEST=GetFeature&srsName=EPSG:4326&TIME=1900-01-01/3000-02-01&outputformat=application/json&TYPENAMES=DSS10-ef59d045-a0cd-4ce0-a25d-5da91e70e965&BBOX=-180,90,180,-90

Please use TYPENAMES=zarr-ef59d045-a0cd-4ce0-a25d-5da91e70e965

It is a zarr collection, that is why

lubojr commented 1 year ago

@lubojr

https://services.sentinel-hub.com/ogc/wfs/sh-id?REQUEST=GetFeature&srsName=EPSG:4326&TIME=1900-01-01/3000-02-01&outputformat=application/json&TYPENAMES=DSS10-ef59d045-a0cd-4ce0-a25d-5da91e70e965&BBOX=-180,90,180,-90

Please use TYPENAMES=zarr-ef59d045-a0cd-4ce0-a25d-5da91e70e965

It is a zarr collection, that is why

This returns <![CDATA[ TYPENAME=zarr-ef59d045-a0cd-4ce0-a25d-5da91e70e965 not found! ]]>

santilland commented 1 year ago

@AlessandroScremin and @dmoglioni the Jupyter Notebook for lake detection without clouds is great, the idea would be to present it at the foss4g workshop, as you won't be able to attend EOX would present it in your name (with clear credit to Rhea and authors). With @aapopescu it was discussed if it would be possible for you to extend the notebook to do some basic area measuring of the lake based on the filtered data and also if possible how the generated result could be added to geodb

Dorothyrono commented 1 year ago

@lubojr

https://services.sentinel-hub.com/ogc/wfs/sh-id?REQUEST=GetFeature&srsName=EPSG:4326&TIME=1900-01-01/3000-02-01&outputformat=application/json&TYPENAMES=zarr-ef59d045-a0cd-4ce0-a25d-5da91e70e965&BBOX=-180,90,180,-90

Also the layers should be availble in the sh -instance id you are using.

if you are using sh-id that ends with ..47c7 it should work now, otherwise let me know which instance id you are using

lubojr commented 1 year ago

@aapopescu

Empty markdown files for Lakes indicators that I am so far aware of are added at:

https://github.com/eurodatacube/eodash/tree/cci-lakes-integration/app/public/data/trilateral

If you do not need some of them (so that two indicators will share the same markdown), then feel free to delete the extra one and I will configure 2 indicators to share a md file.

file name indicator
LWE.md(File) Lake Water Extent timeseries Vanern
LWL.md(File) Lake Water Level timeseries Vanern
Lakes_SWT.md(File) Surface Water Temperature Lakes
Lakes_SWTT.md(File) Surface Water Temperature Time Series
Lakes_S2L2A.md(File) Sentinel 2 L2A for Lake areas
Lakes_WQ_TC.md(File) Water Quality True color Lakes
Lakes_WQ_TURB.md(File) Water Quality Turbidity mean Lakes
Dorothyrono commented 1 year ago

Hi @AlessandroScremin

I have copied the right visualisation for LWQ true color and turbidity where the non lake-pixes are transparent, to your SH account
see [NEW] LWQ - test layers

This is the legend used for turbidity

image

AlessandroScremin commented 1 year ago

@Dorothyrono

thnaks a lot for the support, really appreciated

@lubojr layers in EDC for race have beeen updated accordigly with what @Dorothyrono provide.

Let me know

ALessandro

AlessandroScremin commented 1 year ago

With @aapopescu it was discussed if it would be possible for you to extend the notebook to do some basic area measuring of the lake based on the filtered data and also if possible how the generated result could be added to geodb

hi @santilland and @aapopescu I tried to organize at best this task, but unfortuantely I can not provide the complete workflow as I will be on holiday (already today, but am wotking on it anyway).

What I can provide by Today is an additional set of cells in which looking at the Sentinel 2 acquistion filtered bu full COverage AOI and CC<30%,, I can use the SCL layer and flag the pixels that belong to class 6 (water) and count the number of those pixels, and multiply it by the area of each pixel with pixel size of 100m (and convert in km). I can not go at higher resolution for the limitation of Sentinel hub that allow to use statistical api only for tiles of 2500x2500 pixels. To do that I would need to implement a loop that collect al the tiles, compute area for each of them and then merge the results. As said not feasible for the workshop, but I can provide it using pixel size of 100m, it can be done without other implementation, and the results are quite valid for demonstration purposes...

0 2021-03-05T00:00:00Z 5465.38 1 2021-04-04T00:00:00Z 5197.70 2 2021-04-14T00:00:00Z 5369.04 3 2021-04-19T00:00:00Z 5436.52 4 2021-04-24T00:00:00Z 4035.80 5 2021-05-19T00:00:00Z 5248.66 6 2021-05-29T00:00:00Z 5446.44 7 2021-06-03T00:00:00Z 5390.43 8 2021-06-13T00:00:00Z 5234.46 9 2021-06-28T00:00:00Z 5119.17 10 2021-07-03T00:00:00Z 5320.59 11 2021-07-23T00:00:00Z 4313.01

AlessandroScremin commented 1 year ago

@lubojr @santilland cc: @aapopescu

here the notebook ready with the last changes I proposed above.

kr

Alessandro OpenJN_Retrieve_areas_estimations_over_lakes_foss4g_ready.zip

lubojr commented 1 year ago

Thank you @Dorothyrono for updating the evalscripts and providing the legend of the used colorscale and stretch.

@aapopescu The global 300m cci true color lakes and turbidity lakes have been integrated. https://eodash-trilateral-testing.eox.at/cci-lakes-integration/

AlessandroScremin commented 1 year ago

@lubojr @Dorothyrono

why on sentinel hub with same evalscript I see this

image

and in the dashboard for same date and same evalscrit this?

image

lubojr commented 1 year ago

I will double check.

Edit -> fixed

AlessandroScremin commented 1 year ago

@lubojr @santilland cc: @aapopescu

I found some discerepancy looking at the timeseries and the data available for the lake surface temperature, and there was a misprocessing of the data maps that needed a reprocessing. So I reprocessed all the lakes maps and by the end of the day should be all available. For this I created a new collection ID and new set of dates for the reprocessed data. Also the color scale changed only a bit (only last temperature values changed from 350 to 310 attached hereafter). Here the mapping between the reprocessed collection ID ], dates and the evalscript:

collection ID: LAKES_SURFACE_WATER_TEMPERATURE |  9fdb8c27-9000-4912-b715-1465f840a1db areas: Vanern, Biwa, Balaton, Tonle

avaialbe dates (see files here after)

Vanern_reprocessed_dates.txt Balaton_reprocessed_dates.txt Tonle_reprocessed_dates.txt BIWA_reprocessed_dates.txt

eval script:

''''' //VERSION=3 function setup() { return { input: [{ bands: ["waterTemperature", "dataMask"], // this sets which bands to use }], output: { // this defines the output image type bands: 4, sampleType: "UINT8" } }; }

function evaluatePixel(sample) {

var arr = colorBlend(sample.waterTemperature, [270, 275, 280, 285, 290, 300, 310], [[44,89,150,],[97,154,195],[188,214,230],[246,246,246],[242,200,176],[204,110,90],[146,36,45] ]); if (sample.dataMask==1) arr.push(255); else arr.push(0); return arr; } """

I also added the complete timeseries for the 4 lakes for the Lake surface temperature. it can be found here

database="eodash" table_name = "Lake_Surface_water_Temperature_tri_timeseries"

@santilland