Skip to content

Commit

Permalink
Merge pull request #5132 from nasa-gibs/UAT-v4.32.0
Browse files Browse the repository at this point in the history
UAT-v4.32.0 to Release
  • Loading branch information
christof-wittreich authored Apr 22, 2024
2 parents a0f9760 + c4c110d commit b8f4ebb
Show file tree
Hide file tree
Showing 85 changed files with 12,567 additions and 521 deletions.
Original file line number Diff line number Diff line change
@@ -1,3 +1,7 @@
**NOTE: These Freeze / Thaw (Daily Landscape) images are no longer being updated and may no longer reflect the most recent version of the underlying data set.** MEaSUREs AMSR-E Freeze Thaw data (doi:[10.5067/LJ6SLXNJB2CQ](https://doi.org/10.5067/LJ6SLXNJB2CQ)) are available for download from the NASA NSIDC DAAC without corresponding GIBS/Worldview imagery. Users interested in forward processing freeze/thaw imagery should refer to the <a href="https://worldview.earthdata.nasa.gov/?v=-229.8841449163062,-99.30564687824446,197.5841282463169,102.59903453494508&l=Reference_Labels_15m(hidden),Reference_Features_15m(hidden),Coastlines_15m,SMAP_L3_Passive_Day_Freeze_Thaw(disabled=2),MODIS_Terra_CorrectedReflectance_TrueColor&lg=true">SMAP Freeze/Thaw 36km (L3, Passive, Day)</a> or <a href="https://worldview.earthdata.nasa.gov/?v=-239.19769337486522,-155.03969858850667,175.53006849462886,118.23294210361126&l=Reference_Labels_15m(hidden),Reference_Features_15m(hidden),Coastlines_15m,SMAP_L3_Passive_Enhanced_Day_Freeze_Thaw(disabled=2),MODIS_Terra_CorrectedReflectance_TrueColor&lg=true">SMAP Freeze/Thaw 9km (L3, Passive, Day)</a> layers.

***

The AMSR-E Daily Landscape Freeze/Thaw layer shows the predominant daily frozen or non-frozen (thawed) status of the landscape in vegetated regions where seasonal frozen temperatures are a major constraint to ecosystem processes. It is derived using a temporal change classification of calibrated radiometric brightness temperatures at 36.5 GHz frequency from the AMSR-E instrument. The layer shows 4 states: frozen, non-frozen (thawed), transitional (AM frozen, PM thawed) and inverse transitional (AM thawed, PM frozen) conditions.

The freeze/thaw layer is useful for assessing the impact of freeze/thaw variations on vegetation growing seasons and land-atmosphere carbon exchange; snow cover, permafrost and active layer properties; surface energy and water budgets; distinguishing freeze/thaw dynamics in accordance with regional terrain features, weather events, seasonal and annual climate anomalies, and long-term climate changes.
Expand Down
Original file line number Diff line number Diff line change
@@ -1,27 +1,25 @@
The MODIS Near Real-Time (NRT) Global Flood Product (MCDWD) provides a daily global map of flooding. It is derived from the NRT MODIS Surface Reflectance (MOD09) datasets from both the Terra and Aqua satellites. The Flood Product is available for 3 compositing periods: 1-day, 2-day, and 3-day. For each composite, water detections for all observations (Terra and Aqua) over the compositing period (1, 2, or 3 days) are accumulated, and if the total exceeds the required threshold (1, 2, and 3 observations, respectively), the pixel is marked as water. (Note: 1-day product not yet available in Worldview)
The MODIS Near Real-Time (NRT) Global Flood Product (MCDWD) provides a daily global map of flooding. It is derived from the NRT MODIS Surface Reflectance (MOD09) datasets from both the Terra and Aqua satellites. The Flood Product is available for 3 compositing periods: 1-day, 2-day, and 3-day. For each composite, water detections for all observations (Terra and Aqua) over the compositing period (1, 2, or 3 days) are accumulated, and if the total exceeds the required threshold, the pixel is marked as water. (Note: 1-day product not yet available in Worldview)

Users are advised to compare the flood product against the contributing MODIS reflectance imagery (such as 7-2-1 Corrected Reflectance; search for “721” after clicking “Add Layers”) , for the compositing period to ensure reported flood areas do not correspond to areas of cloud shadow. [Learn more...](https://earthdata.nasa.gov/earth-observation-data/near-real-time/mcdwd-nrt#ed-flood-faq)

As of January 12, 2023, a topographic filter has been applied to remove water detections from mountainous areas, greatly reducing the number of terrain shadow false-positives in such areas. These areas appear in all products as "Insufficient Data" (gray in default Worldview display).

#### Current Issues
- The flood layers are displaying a large number of flood pixels in the far north at present (generally above 60N). Most of these are false-positive detections, a result of the large number of overlapping images towards the poles in the source MODIS imagery (due to the nature of the spacecraft's polar orbit). During summer when such regions have long periods of daylight, even more observations are available, but this increases the chance that repeated cloud-shadow false-positives pass the compositing requirement, and contaminate the product. We are working on various adjustments to the algorithm to minimize this.

At the same time, note there is actual major flooding in the region along various Arctic rivers (as of mid June 2022). These are easily distinguishable by the spatial pattern of flooding (occurring clearly along river courses). The false positives are generally scattered around randomly, or appear similar to cloud spatial patterns (not conforming to local hydrology).

- Far west tiles (Hawaii, Alaska): Due to issues with processing imagery around the international dateline for this product, far west tiles will sometimes appear with data at the start of the day, long before Terra or Aqua have observed for the day. Users are advised to disregard such data, until the Corrected Reflectance layers confirm current-day observations have been processed.

#### Limitations
Prior to introduction of product Release 1 on April 16, 2024 the flood layers displayed a large number of flood pixels in the far north (generally above 60N latitude). Most of these are false-positive detections, resulting from the large number of overlapping images towards the poles in the source MODIS imagery (due to the nature of the spacecraft's polar orbit). Release 1 includes new compositing requirements that has greatly reduced, but not entirely eliminated, these false-positives. The archive product imagery available in Worldview until April 11, 2024 continues to contain these errors. Users are advised to consider the expected patterns of flooding, in order to more readily discern false-positives; real flooding is generally easily distinguishable by its spatial pattern (usually occurring along river courses or adjacent to existing water bodies). Whereas the false positives are generally scattered around randomly, or appear similar to cloud spatial patterns (not conforming to local hydrology).

Common situations in which the flood product may be unable to accurately identify flood include:

- Surface obscuration: clouds and canopy cover can block view of water on the surface. Buildings can also provide a “dry” roof, diluting the signal from surrounding water.
- Cloud shadow false-positives: cloud shadows are detected as water by the algorithm; when they recur in the same location over the compositing period, false positives are likely to be reported. Longer compositing periods help minimize this. Please check reflectance imagery of dates contributing to composite to rule these out, if reported flood looks unusual or suspicious.
- Terrain shadow false-positives: terrain shadows can create false-positives in mountains, generally only in wintertime. These are typically easy to identify due to their pattern (reflecting topography), and by comparison to reflectance imagery.
- Terrain shadow false-positives: terrain shadows can create false-positives in mountains, generally only in wintertime. These are typically easy to identify due to their pattern (reflecting topography), and by comparison to reflectance imagery. As of January 12, 2023, a topographic filter has been applied which greatly reduces terrain shadow false-positives.
- Dark volcanic rock or soils: such areas can be identified as water, and thus will routinely be reported as flood.
- Springtime snow melt ponding on fields: such water can appear as pixellated flood across flat areas of agricultural fields. Although this is unusual water, it is often very shallow, and not moving, and thus typically not a flood in the normal sense. Checking the reflectance imagery will typically show such areas on the edge of larger areas of snow extent, or, looking back in time, will show them recently covered by snow.
- Springtime snow melt ponding on fields: such water can appear as pixelated flood across flat areas of agricultural fields. Although this is unusual water, it is often very shallow, and not moving, and thus typically not a flood in the normal sense. Checking the reflectance imagery will typically show such areas on the edge of larger areas of snow extent, or, looking back in time, will show them recently covered by snow.

#### Spatial Coverage
Non-polar global land areas (below 70 degrees latitude), comprising 223 10x10 degree tiles (see Figure 4 in [User Guide](https://www.earthdata.nasa.gov/s3fs-public/imported/MCDWD_UserGuide_RevB.pdf) for included tiles).
Non-polar global land areas (below 80 degrees latitude), comprising 287 10x10 degree tiles (see [product homepage](https://www.earthdata.nasa.gov/learn/find-data/near-real-time/modis-nrt-global-flood-product) for map of included tiles).

#### Sensor/Image Resolution
Nominal equatorial resolution is ~232 m per pixel, with resolution increasing toward the poles (~116 m at 60 degrees latitude). Note the higher apparent resolution towards the poles is simply an artifact of the lat/lon (geographic) projection used, and not intrinsic to the data.
Expand All @@ -33,4 +31,4 @@ To help estimate if the final flood product (for the day) is available in Worldv

Flood products displayed in Worldview are updated every 30 minutes, approximately on the hour and at 30 minutes past the hour.

References: MCDWD_L3_NRT [doi:10.5067/MODIS/MCDWD_L3_NRT.061](https://doi.org/10.5067/MODIS/MCDWD_L3_NRT.061)
References: MCDWD_L3_NRT [doi:10.5067/MODIS/MCDWD_L3_NRT.061](https://doi.org/10.5067/MODIS/MCDWD_L3_NRT.061)
Original file line number Diff line number Diff line change
@@ -1,27 +1,25 @@
The MODIS Near Real-Time (NRT) Global Flood Product (MCDWD) provides a daily global map of flooding. It is derived from the NRT MODIS Surface Reflectance (MOD09) datasets from both the Terra and Aqua satellites. The Flood Product is available for 3 compositing periods: 1-day, 2-day, and 3-day. For each composite, water detections for all observations (Terra and Aqua) over the compositing period (1, 2, or 3 days) are accumulated, and if the total exceeds the required threshold (1, 2, and 3 observations, respectively), the pixel is marked as water. (Note: 1-day product not yet available in Worldview)
The MODIS Near Real-Time (NRT) Global Flood Product (MCDWD) provides a daily global map of flooding. It is derived from the NRT MODIS Surface Reflectance (MOD09) datasets from both the Terra and Aqua satellites. The Flood Product is available for 3 compositing periods: 1-day, 2-day, and 3-day. For each composite, water detections for all observations (Terra and Aqua) over the compositing period (1, 2, or 3 days) are accumulated, and if the total exceeds the required threshold, the pixel is marked as water. (Note: 1-day product not yet available in Worldview)

Users are advised to compare the flood product against the contributing MODIS reflectance imagery (such as 7-2-1 Corrected Reflectance; search for “721” after clicking “Add Layers”) , for the compositing period to ensure reported flood areas do not correspond to areas of cloud shadow. [Learn more...](https://earthdata.nasa.gov/earth-observation-data/near-real-time/mcdwd-nrt#ed-flood-faq)

As of January 12, 2023, a topographic filter has been applied to remove water detections from mountainous areas, greatly reducing the number of terrain shadow false-positives in such areas. These areas appear in all products as "Insufficient Data" (gray in default Worldview display).

#### Current Issues
- The flood layers are displaying a large number of flood pixels in the far north at present (generally above 60N). Most of these are false-positive detections, a result of the large number of overlapping images towards the poles in the source MODIS imagery (due to the nature of the spacecraft's polar orbit). During summer when such regions have long periods of daylight, even more observations are available, but this increases the chance that repeated cloud-shadow false-positives pass the compositing requirement, and contaminate the product. We are working on various adjustments to the algorithm to minimize this.

At the same time, note there is actual major flooding in the region along various Arctic rivers (as of mid June 2022). These are easily distinguishable by the spatial pattern of flooding (occurring clearly along river courses). The false positives are generally scattered around randomly, or appear similar to cloud spatial patterns (not conforming to local hydrology).

- Far west tiles (Hawaii, Alaska): Due to issues with processing imagery around the international dateline for this product, far west tiles will sometimes appear with data at the start of the day, long before Terra or Aqua have observed for the day. Users are advised to disregard such data, until the Corrected Reflectance layers confirm current-day observations have been processed.

#### Limitations
Prior to introduction of product Release 1 on April 16, 2024 the flood layers displayed a large number of flood pixels in the far north (generally above 60N latitude). Most of these are false-positive detections, resulting from the large number of overlapping images towards the poles in the source MODIS imagery (due to the nature of the spacecraft's polar orbit). Release 1 includes new compositing requirements that has greatly reduced, but not entirely eliminated, these false-positives. The archive product imagery available in Worldview until April 11, 2024 continues to contain these errors. Users are advised to consider the expected patterns of flooding, in order to more readily discern false-positives; real flooding is generally easily distinguishable by its spatial pattern (usually occurring along river courses or adjacent to existing water bodies). Whereas the false positives are generally scattered around randomly, or appear similar to cloud spatial patterns (not conforming to local hydrology).

Common situations in which the flood product may be unable to accurately identify flood include:

- Surface obscuration: clouds and canopy cover can block view of water on the surface. Buildings can also provide a “dry” roof, diluting the signal from surrounding water.
- Cloud shadow false-positives: cloud shadows are detected as water by the algorithm; when they recur in the same location over the compositing period, false positives are likely to be reported. Longer compositing periods help minimize this. Please check reflectance imagery of dates contributing to composite to rule these out, if reported flood looks unusual or suspicious.
- Terrain shadow false-positives: terrain shadows can create false-positives in mountains, generally only in wintertime. These are typically easy to identify due to their pattern (reflecting topography), and by comparison to reflectance imagery.
- Terrain shadow false-positives: terrain shadows can create false-positives in mountains, generally only in wintertime. These are typically easy to identify due to their pattern (reflecting topography), and by comparison to reflectance imagery. As of January 12, 2023, a topographic filter has been applied which greatly reduces terrain shadow false-positives.
- Dark volcanic rock or soils: such areas can be identified as water, and thus will routinely be reported as flood.
- Springtime snow melt ponding on fields: such water can appear as pixellated flood across flat areas of agricultural fields. Although this is unusual water, it is often very shallow, and not moving, and thus typically not a flood in the normal sense. Checking the reflectance imagery will typically show such areas on the edge of larger areas of snow extent, or, looking back in time, will show them recently covered by snow.
- Springtime snow melt ponding on fields: such water can appear as pixelated flood across flat areas of agricultural fields. Although this is unusual water, it is often very shallow, and not moving, and thus typically not a flood in the normal sense. Checking the reflectance imagery will typically show such areas on the edge of larger areas of snow extent, or, looking back in time, will show them recently covered by snow.

#### Spatial Coverage
Non-polar global land areas (below 70 degrees latitude), comprising 223 10x10 degree tiles (see Figure 4 in [User Guide](https://www.earthdata.nasa.gov/s3fs-public/imported/MCDWD_UserGuide_RevB.pdf) for included tiles).
Non-polar global land areas (below 80 degrees latitude), comprising 287 10x10 degree tiles (see [product homepage](https://www.earthdata.nasa.gov/learn/find-data/near-real-time/modis-nrt-global-flood-product) for map of included tiles).

#### Sensor/Image Resolution
Nominal equatorial resolution is ~232 m per pixel, with resolution increasing toward the poles (~116 m at 60 degrees latitude). Note the higher apparent resolution towards the poles is simply an artifact of the lat/lon (geographic) projection used, and not intrinsic to the data.
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@

The dynamically generated Reflectance (Bands 5-4-3, False Color) imagery layer is a false color composite that is used to gauge plant health and assess plant density. Plants reflect in near infrared and green light, while absorbing in red. Areas with better vegetation health appear red, and denser plant growth is a darker red. Cities, urban areas, and exposed ground appear grey or tan and water is blue or black.

The Reflectance imagery layer from Landsat 8 and 9/OLI product (L30) is available through the HLS project from the Operational Land Imager (OLI) aboard the Landsat 8 and 9 satellites. The sensor resolution is 30 m, imagery resolution is 30 m, and the temporal resolution is daily with an 8 day revisit time. The imagery is available in Worldview/GIBS approximately 2 - 4 days after satellite overpass. There is a separate combined Sentinel 2A and Sentinel 2B imagery layer available.
The Reflectance imagery layer from Landsat 8 and 9/OLI product (L30) is available through the HLS project from the Operational Land Imager (OLI) aboard the Landsat 8 and 9 satellites. The sensor resolution is 30 m, imagery resolution is 30 m, and the temporal resolution is daily with an 8 day revisit time. The imagery is available in Worldview/GIBS approximately 2 - 4 days after satellite overpass. There is a separate combined Sentinel-2A and Sentinel-2B imagery layer available.

This imagery layer is provided dynamically through the [NASA Interagency Implementation and Advanced Concepts Team (IMPACT)](https://earthdata.nasa.gov/esds/impact). As it is dynamically generated, it may take slightly longer to display than normal. The imagery is also only available at higher zoom levels.

Expand Down
Loading

0 comments on commit b8f4ebb

Please sign in to comment.