Evaluating State-of-the-Art Climate Datasets

New paper alert! A fully coupled climate reanalysis by Vince Cooper covering 1850-2023. We used strongly coupled data assimilation on observations of sea surface temperature, land-based air temperature, sea-level pressure over the ocean, and satellite sea-ice concentration at monthly resolution. As far as we know, this is the first time that these fields have been simultaneously reconstructed over the historical period. Results show significant low-frequency variance in ENSO, with a peak near the start of the 20th century, muted modern cooling trends in Southern Ocean SST (see figure below), a decline in Arctic sea-ice area since the 19th century, and relatively small changes in Antarctic sea-ice area. Additional key points: * Most reanalysis datasets consider each component of the climate system independently (i.e., separate atmospheric and oceanic reanalyses), leading to inconsistencies in coupled variability. Here, we use strongly coupled data assimilation, which means that all observations update every component of the climate system. * Efficient emulators are used to propagate the memory of past observations forward in time. We use cyclostationary linear inverse models trained on 8 CMIP6 model simulations to include the role of model error in the reconstructions. These models are used to create 8 separate reanalyses, propagating the full error covariance matrix for all climate variables. * A 1600-member ensemble is created by sampling the posterior distributions in a dynamically consistent process, providing a large sample of equally likely reanalyses of historical climate. This provides a rich dataset for exploring climate variability with uncertainty quantification. The preprint can be found here: https://lnkd.in/gbEtR4Jw

In 2022, I was part of a research team that took on a critical challenge: 𝐡𝐨𝐰 𝐝𝐨 𝐲𝐨𝐮 𝐚𝐬𝐬𝐞𝐬𝐬 𝐫𝐚𝐢𝐧𝐟𝐚𝐥𝐥 𝐢𝐧 𝐫𝐞𝐠𝐢𝐨𝐧𝐬 𝐰𝐡𝐞𝐫𝐞 𝐫𝐚𝐢𝐧 𝐠𝐚𝐮𝐠𝐞𝐬 𝐚𝐫𝐞 𝐬𝐜𝐚𝐫𝐜𝐞? Our focus was Sudan, a country where climate-driven decisions are crucial, but data gaps make it difficult to plan and prepare. We turned to satellite and gridded rainfall datasets, tools that have become essential in modern hydroclimatic research. What we found was both promising and eye-opening. 📍 While most rainfall products showed a tendency to underestimate rainfall, especially on annual and monthly scales, two stood out: 𝐂𝐇𝐈𝐑𝐏𝐒 𝐚𝐧𝐝 𝐂𝐑𝐔 𝐝𝐚𝐭𝐚𝐬𝐞𝐭𝐬 consistently performed best, especially in the western and southern regions of Sudan. 🌧️ We discovered that summer rainfall (the main rainy season) is captured more accurately than annual totals, especially in mountainous areas. And when we explored deeper, we noticed a significant link between rainfall trends and the Atlantic Multidecadal Oscillation (AMO), with some regions showing correlations as high as 90%. This was more than just data and numbers; it was a reflection of how remote sensing, when used wisely, can support 𝐜𝐥𝐢𝐦𝐚𝐭𝐞 𝐫𝐞𝐬𝐢𝐥𝐢𝐞𝐧𝐜𝐞, 𝐟𝐨𝐨𝐝 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐲, 𝐚𝐧𝐝 𝐫𝐢𝐬𝐤 𝐩𝐫𝐞𝐩𝐚𝐫𝐞𝐝𝐧𝐞𝐬𝐬 in the very places that need it most. 🔍 The study underscored the value of CHIRPS data for monitoring rainfall variability and extreme events, and why it should be a go-to resource for decision-makers in data-scarce environments. Being part of this work reminded me of the power of science and collaboration in driving evidence-based action. It’s research like this that fuels real-world solutions, and I’m proud to have contributed to it.  Richard Anyah Mohamed Abdallah Ahmed Alriah, PhD Aslak Grinsted Hans Hersbach Lorenz Ewers Victor Ongoma Nixon Mutai #ClimateResearch #RainfallData #RemoteSensing #Sudan #ClimateResilience #CHIRPS #Hydroclimatology #SatelliteData #ClimateScience #DataForDevelopment

Analysis-Ready, Cloud Optimized ERA5 -- https://lnkd.in/gP_Y8YCC <-- git hub repository -- H/T Robbi Bishop-Taylor [I am trying to understand all the technical details and use case(s) for this project, but I will get there – but thought others might find it of interest] “Exploring the mind-blowing ARCO-ERA5 dataset: hourly data for almost 300 climate variables, available globally from 1940 onwards!🤯 All packaged into a single cloud-friendly Zarr file, and loadable with a single line of Python code - crazy!..” -- “[Their] goal is to make a global history of the climate highly accessible in the cloud. To that end, [they] present a curated copy of the ERA5 corpus in Google Cloud Public Datasets… ERA5 is the fifth generation of ECMWF's Atmospheric Reanalysis. It spans atmospheric, land, and ocean variables. ERA5 is an hourly dataset with global coverage at 30km resolution (~0.28° x 0.28°), ranging from 1979 to the present. The total ERA5 dataset is about 5 petabytes in size… A reanalysis is the "most complete picture currently possible of past weather and climate." Reanalyses are created from assimilation of a wide range of data sources via numerical weather prediction (NWP) models… So far, [they] have ingested meteorologically valuable variables for the land and atmosphere. From this, [they] have produced a cloud-optimized version of ERA5, in which [they] have converted grib data to Zarr with no other modifications. In addition, [they] have created "analysis-ready" versions on regular lat-lon grids, oriented towards common research & ML workflows. This two-pronged approach for the data serves different user needs. Some researchers need full control over the interpolation of data for their analysis. Most will want a batteries-included dataset, where standard pre-processing and chunk optimization is already applied. In general, [they] ensure that every step in this pipeline is open and reproducible, to provide transparency in the provenance of all data…” #GIS #spatial #mapping #remotesensing #ARCO #ERA5 #global #spatialanalysis #spatiotemporal #code #model #modeling #visualisation #global #GitHub #opensource #hourly #GoogleCloudPublicDatasets #climate #climatechange #ECMWF #atmosphere #weather #extremeweather #NWP #weatherprediction #numericalweatherprediction #meteorology #interpolation #preprocessing #dataprovenance 

🌍 Exciting news for climate researchers and enthusiasts! 🌦️ The ECMWF ERA5 reanalysis dataset is revolutionizing our understanding of global climate patterns. 🔄 ERA5, the fifth generation of ECMWF atmospheric reanalysis, combines cutting-edge model data with observations from around the world, creating a comprehensive and consistent dataset. It's a significant upgrade from its predecessor, ERA-Interim reanalysis. One of the remarkable offerings of ERA5 is the ERA5 DAILY dataset, which provides aggregated daily values for seven key climate parameters, including 2m air temperature, total precipitation, and wind components. Daily aggregates such as mean sea level pressure and surface pressure offer valuable insights into daily weather patterns. For researchers and data enthusiasts, ERA5 DAILY opens up avenues for exploring climate trends and understanding weather phenomena on a global scale. From tracking changes in precipitation patterns to studying wind dynamics, the possibilities are endless. 📊 And here's where the magic happens: utilizing tools like Google Earth Engine, we can harness the power of ERA5 data for localized analysis and visualization. Check out this code snippet using ERA5 DAILY to analyze precipitation patterns in the Ceará region of Brazil! See code bellow: // Defining the region of interest var gaul1 = ee.FeatureCollection("FAO/GAUL/2015/level1"); var brazilStates = gaul1.filter(ee.Filter.eq('ADM0_NAME', 'Brazil')); var roi = brazilStates.filter(ee.Filter.eq('ADM1_NAME', 'Ceara')); // Setting the study area Map.centerObject(roi); Map.addLayer(roi); // Setting the time interval var starting = '2010-01-01'; var ending = '2023-01-01'; // Applying unit conversion var eraPrec = ee.ImageCollection("ECMWF/ERA5_LAND/DAILY_AGGR") .filterDate(starting, ending) .filterBounds(roi); // Printing the collection print('Collection:', eraPrec); print('Number of images:', eraPrec.size()); // Function to convert m to mm and add property to the collection var Precipitation = function(img){ // Precipitation units are depth in meters: divide to get m / mm var bands = img.select('total_precipitation_sum').multiply(1000).clip(roi); return bands.rename('total_precipitation_sum') .set('date', img.date().format('YYYY-MM-dd')) .copyProperties(img,['system:time_start','system:time_end']); }; var eraPrecConverted = eraPrec.map(Precipitation); rest of the code down in the comments #ERA5 #ClimateData #ClimateResearch #DataScience #ECMWF #EarthObservation #ClimateChange #WeatherPatterns #GoogleEarthEngine #DataVisualization #Copernicus #ClimateAction #javascript #codetutorial #remotesensing