Land cover polygons and files in the WebGIS

Until today the manual land-use editor could only be painted with the brush, stroke after stroke: covering a whole field or a patch of woodland meant going over the area again and again until it was fully coloured, with the risk of leaving holes or spilling past the boundary. Alongside the brush, the eraser and the hand there is now a polygon, and it works exactly like the area drawing you already know elsewhere on the platform: pick the cover, click the vertices of the area, close the shape and the inside is filled all at once with that cover's colour. While you draw you see the rubber-band perimeter following the mouse, the vertices already placed and the surface area computed live, so you know straight away how much you are covering. You close the shape by clicking the first vertex again or with a double click, and if you have placed fewer than three points the platform will not let you close an area that does not exist. The tool's bar carries three dedicated commands: finish the shape, remove the last point if you missed your aim, cancel everything and start over. Every completed polygon counts as a single step in the history, so one "undo" removes the whole filled area, exactly as it would a single brush stroke. While the tool is armed the map cannot be dragged — otherwise every attempt to move would leave an unwanted vertex behind — but zooming keeps working with the wheel and with the buttons, so you can move in close to place a vertex precisely. The Esc key cancels only the shape you are tracing and leaves the tool in your hand, rather than throwing you out of the drawing session. We also moved the polygon bar below the top row of controls, because it used to sit under the basemap selector and the "Cancel" button was simply not clickable. You will find the new tool at XNatura → Monitoring → Site → Land use → Manual editor, in the tool column on the left.

The WebGIS panel on the left only ever showed layers, that is the result of what you had uploaded, and there was no way back to the source file: if you had imported three different exports in the same afternoon, working out which layer came from which file was an exercise in memory. That panel now has two tabs, Layers and Files, and the second one lists every file uploaded to the site with its original name, its format, the upload date, who uploaded it and how many geometries it contains. From each row you can frame the file on the map with one click, so you immediately see where something you imported months ago actually sits. You can also download it again: the platform re-exports it in the format it was uploaded in, which is handy when the original file is no longer on your computer or when you have to hand it to a colleague. In the same pass we fixed a flaw you noticed right after drawing: a layer you have just created by hand has no file behind it yet, and it used to keep its loading spinner on forever, as if it were downloading something that would never arrive. Those layers now simply present themselves as ready, with no phantom waiting. We also removed a pointless reload: a layer's geographic content was re-downloaded every time the list refreshed, even when nothing had changed, and on heavy files you could feel it. It is now re-read only when the layer has genuinely changed, so the map responds faster and uses less data. You will find the new tab at XNatura → Monitoring → Site → WebGIS, at the top of the layers panel.

The Scenarios page compares the current state of the land with the transformations you are simulating, row by row and indicator by indicator. Until today that table said nothing about how vigorous the vegetation is, which is one of the first things you want to know when assessing an intervention: you could read naturalness, imperviousness, surface areas and the biodiversity indicators, but not how green and thriving the resulting area would be. There is now a column dedicated to mean NDVI, the index that measures from satellite imagery how dense and healthy the vegetation is. Because the table and the "Scenario evolution" chart share the same columns, the indicator appears in both at once, with nothing extra to switch on in either place. We also fixed the way values are printed: every number in the table used to be rounded to a whole figure, because every other indicator here is a percentage or a count, so a mean NDVI of 0.40 came out as "0" and a difference of 0.05 between two scenarios became "+0". Each column now declares how many decimals it needs, and NDVI shows two, so the differences between one scenario and another can actually be read. The column only appears on platforms that have NDVI among their active indicators, so it does not add an empty column for anyone who does not use it. Remember that NDVI runs between −1 and 1: high values mean dense vegetation in good condition, low values point to bare soil, built surfaces or vegetation under stress. You will find the new column at XNatura → Monitoring → Site → Scenarios, in the comparison table.

When a sensor stops sending data the first question is always the same: is it broken, or is it simply off the network? Until today that answer was not readable in the platform and required a manual check, with all the waiting that entails. In the census sensor management panel, administrators now find a box that queries the connectivity of the device's data card and reports the mobile operator, the card's status, when it last connected and how many megabytes it has used in total and in the current month. With that information you can tell a real failure from a plain lack of coverage or a suspended card within seconds, and avoid scheduling a site visit that was never needed. The check only starts when you press the button, not when the panel opens: scrolling through many sensors one after another, an automatic request for each one would burn through the allowed number of queries in moments, and at that point nobody would get an answer at all. The result stays valid for ten minutes, so reopening the same device shortly afterwards shows the figure straight away without querying again. The box covers seven different situations, including the case where the card answers but the serial number on record does not match, and the case where the last connection reported comes from the operator's own records rather than a real session and could therefore be months old: in both cases it tells you so explicitly, rather than showing you a figure that looks trustworthy and is not. The block is read-only, appears only to administrators and only when the surveyed point really does have a device attached. You will find it at XNatura → Monitoring → Census → Sensors, by opening a sensor's card.

The storytelling copy on your platform — the opening title, the impact cards, the stories, the captions — is written straight on the page, by clicking the text and rewriting it. It was, however, completely flat prose: there was no way to highlight a figure, the name of a project or the one sentence you want the reader to remember. Now, when you select a portion of text while you are writing, a small bar appears with bold, italic and a command to clear the formatting. The keyboard shortcuts you are used to work too, Cmd+B and Cmd+I on Mac or Ctrl+B and Ctrl+I on Windows, and the buttons light up to tell you whether the selected text is already bold or italic. The formatting is shown correctly to page visitors as well, in every section where this copy appears: the cover, the introduction section, the impact cards with their lightbox, and the stories, on both desktop and phone. The character counter keeps counting the text exactly as it is saved and now shows the visible length next to it, so you always know how much room you have left against the limit. If you genuinely need an asterisk in the text, simply put a backslash in front of it and it will stay as it is, without being read as a formatting command. We paid particular attention to never letting visitors see the formatting symbols instead of the formatted text, and that applies to image alternative descriptions and to downloaded file names too. You will find the new bar at XNatura → Overview → Storytelling, by clicking an editable text and selecting part of it.

We were told that the background of the project chapter was so blurred that the photograph you had uploaded was unrecognisable: you could sense an image, but not tell which one. The blur was a fixed value decided once and for all, identical for every photo, and with detail-rich images the result was a smear of colour. The default blur is now much lighter and, more importantly, it has become your setting: inside the "Change background" dialog there is a slider you can raise or lower until you find the right balance between the legibility of the text above and the recognisability of the photo below. We also lightened the dark veil covering the middle band of the image, which is exactly where the photograph shows most, while leaving the top and bottom edges untouched, since that is where the title and the last paragraph sit and where text contrast has to be protected. The second report was about uploading: it was not clear what kind of image you needed in order to get a good-looking cover. The upload dialog now states the recommended aspect ratio, the recommended pixel size and the minimum acceptable width up front, and during cropping it tells you what will be cut and at what size the selection will be exported. If the region you selected is smaller than the export size it warns you that it will be scaled up and will therefore look soft, so you can fix the crop before saving rather than discovering it once the page is live. Finally we widened the space available to the cover title, which used to wrap onto four very narrow lines with a long title, and increased the bleed at the edges of the image, because a very high blur left a thin light halo along all four sides. You will find all of it at XNatura → Overview → Storytelling, by opening "Change background" on the project chapter.

The card counting monitored sites had disappeared from the "The numbers that really matter" section of the Overview, and with it the only place on the page that stated how many sites the platform is following. The result was disorienting: the Overview no longer mentioned sites at all, while the Monitoring page kept showing their total, and the two figures could no longer be compared at a glance. The card had been removed deliberately some time ago for a different request, then a later report asked for it to be put back: we have therefore restored it exactly as it was, in its original position between adopted oases and sensors. The value is once again the site count, with the total surface in hectares underneath, rounded to one decimal as before. The card only appears when there is at least one site, so you will never see a meaningless zero, and the hectares line only appears when the surface is visibly greater than zero at the precision shown. It stays tied to the Monitoring toggle in the Overview settings: if you choose not to show that part, the card keeps hiding along with the rest. To stop it being deleted a third time we left an internal note reconstructing both requests and the final decision, so this card's history does not get lost. You need to do nothing to see it again: open XNatura → Overview → The numbers that matter and it will be in its place.

On the Overview the adoption cards and the supporters map told two different stories: a card could state two adopted hives while the map just below wrote "Adopted hives 0" and the matching filter returned nothing at all. The cause is that the two sections read counts that are built differently — one cumulative, the other tied to the adoptions active at this exact moment — and no page setting could make them agree. We picked the most honest route: the map takes the cards' figures, which are the more complete ones, and when some adoptions have no position to draw it says so explicitly instead of pretending they do not exist. That way the number you read is the same across the whole page and, if a point is missing from the map, you know why. In the same pass we removed the filters that could never work: a filter pill with a zero next to it can only empty the list, so on the territory map and on the supporters map the entries with a zero count are no longer drawn at all. If a change of period drops the category you had selected to zero, the selection automatically returns to "All", so you are not stranded in front of an empty list filtered by a pill you can no longer see. When no category survives, the whole filter row disappears, because a lone "All" pill filters nothing and only takes up space. Counting, ordering, colours and search are unchanged, and the same correction applies to the phone versions of both maps. You will find them at XNatura → Overview → The map of your sites.

With the period selector set to "Ongoing", the Regeneration section of the Overview still showed an amount of absorbed CO2, even on platforms where the planned plants had not yet been put in the ground and would only arrive in the field months later. The figure was not made up, but it answered a different question from the one you were asking: the sources the absorption is computed from reason in whole years and cannot narrow the count to a window of a few days or weeks, so a period ending in summer was also credited with absorption from plants due to be planted in autumn. For anyone opening the Overview precisely to see where the project stands today, that is a substantial difference. Absorption is now only shown if at least one plant is genuinely alive within the selected period, because the CO2 is absorbed by those plants and cannot exist before them. The plant count, unlike absorption, is already correctly tied to the period and flips exactly on the day planting begins: it is therefore the guard for the CO2 figure. Select a period that includes the autumn or the whole year and you will see the values come back, as they should. Nothing changes in the way absorption is calculated: what changes is that it is no longer credited to months when those plants were not there. You will find the section at XNatura → Overview → Regeneration, using the period selector at the top.

At the bottom of the Overview there is a section gathering the in-depth blocks, and its title was fixed copy: it announced "Details and reporting" even when the reporting part had been switched off and only one block remained below, about something else entirely. Visitors were therefore reading a heading that promised content that was not there, which made the section look incomplete or broken. Title and subtitle are now derived from the blocks the visitor actually sees, with four possible combinations: both blocks present, reporting only, the project block only, or — a case that concerns administrators alone — neither of the two. The title used when both blocks are present has been reworded so that it suits both of them, not just reporting. Administrators keep seeing every block, since they are the ones managing the toggles, but they read the same heading their clients read, so they always know how the page looks from the outside. Which blocks render and the way changes are saved are unchanged: only the title introducing them changes. It is a correction that needs nothing from you and that shows up mostly on platforms keeping only part of the section active. You can see it at the bottom of XNatura → Overview.

The platform's storytelling page is built out of thematic blocks, and two of them were showing content even when they had nothing to tell. The first is the social impact block: it always rendered, and on a platform with no adopted schools it showed the label, the heading, the line "0 indicators" and an invitation to adopt a school — a whole block built around a zero. That block now appears only when there is at least one adopted school, exactly as the monitoring and regeneration blocks already did, and it relies on the same count the technical overview uses to decide whether that section is active. The second is the map legend in the chapter about the territory: it always listed both categories of point, even on platforms that only have one, so on a platform with monitored sites alone the legend announced a category the map never drew. Worse, the note accompanying one of the entries mentioned hives even where there are no hives. The legend is now built from the very same point lists that generate the markers on the map, so the two can no longer tell different stories: an entry appears only if at least one point of that type exists, and the note about hives only if the sites really do have some. While the points are still loading nothing is hidden, so an entry cannot vanish and reappear in front of the reader. The legend container stays in place even when there are zero entries, so the rest of the page does not shift. You will find the storytelling page at XNatura → Overview → Storytelling.

The Risks and Opportunities page opens with a few summary cards: how many sites you have, how much economic value is exposed, how many sit in the high or extreme band. Those counts also included the sites for which the risk assessment was never generated, that is sites with no point on the map and no presence in any chart on the page. The result was a page contradicting itself: the card at the top declared two sites while the doughnut chart below represented only one, and there was no way to tell which of the two figures was right. Every summary on the page now uses the same criterion already applied to decide whether there is any data to show, that is it counts only the sites with an assessment actually available. Site count, economic total and the number of sites in the high or extreme band therefore describe the same set, and what you read at the top matches what you see in the charts. The sites still to be assessed do not disappear: they stay listed in the detail tables, dimmed, exactly as before, so you know which assessments are still missing. The check that decides whether the page has enough data to be shown has also been returned to its original behaviour, because a stricter version could have emptied the whole page in certain cases. You will find the cards at the top of XNatura → Monitoring → Risks and Opportunities.

The cadastral layer, which overlays the map to show parcel boundaries, was often reported to us as "it does not load properly". There were two reasons, both to do with how the public service supplying those images behaves, and both something we could address from our side. The first: when the service is under load it does not answer some of the tiles the map is made of, and the result was a silent hole in the mosaic, with no warning at all. Each missing tile is now requested again up to three times, with progressively longer waits between attempts, so in the great majority of cases the hole closes by itself before you even notice it. The second: below a certain zoom level the service answers correctly but with a completely blank image, so the layer simply looked broken. Now, when you are too far out for the cadastre to be drawn, a notice appears telling you so and inviting you to zoom in, and requests that would certainly come back blank are not even sent. We also reduced the constant churn of tiles while zooming: every intermediate step used to start and then abort a full set of images, wasting time and bandwidth, whereas tiles already downloaded are now kept for longer. Overall the layer appears faster, more complete and with fewer pointless waits. You switch it on at XNatura → Monitoring → Map → Cadastre.

Among the available basemaps there is "History", which lets you see the land as it looked in satellite surveys from past years and scroll back and forth through time. It was, however, a feature that tested your patience: the details of each historical release were requested one at a time, in a queue, and every request also had to wake up a service that had fallen asleep in the meantime, for a total wait that in the worst cases exceeded three minutes before the timeline appeared. On top of that nothing was kept, so switching to another basemap and back meant paying the whole bill again. Requests now go out in groups rather than one at a time, and results are kept aside for the whole session: in our measurements the cold wait dropped from over three minutes to about one minute, and the warm one from several seconds to less than one. The background image no longer waits for all the details to arrive either: it appears straight away with the most recent release, and the timeline fills in while you are already looking at the map. We also removed a particularly annoying behaviour where moving the map during loading restarted everything from scratch — which happened regularly, because automatically framing the site produces several movements in a row. If something goes wrong you can now see it: instead of an empty, unexplained date selector, a real error message appears. Try it at XNatura → Monitoring → Map, choosing "History" among the basemaps.

The "Sensors in the field" card on the Overview has a question mark next to it explaining exactly what that number counts. That explanation was a single compact paragraph in which three different ideas — which sensors you actually have installed, what kind of data they collect and what the count includes — chased each other with no line breaks at all, because the little panel could not represent line breaks and turned them into plain spaces. The text is now organised into three blocks separated by white space: first the real make-up of your sensor fleet, then the sentence about the data collected, then what falls inside the count and the pointer to the monitoring section where you find the detail. The number and the make-up are not an example: they are derived from the sensors genuinely present on your sites, model by model, so every platform reads its own. Singular and plural follow the number too, and a platform with no sensors installed yet reads a sentence written for that case instead of a zero count followed by an empty bracket. We also took care of accessibility: the same text is used as the description of the help button for anyone navigating with a screen reader, and there the block structure makes no sense, so in that context it is flattened into a single continuous sentence that listens well. Automatic linking of web addresses inside the explanations keeps working as before. The change applies to both the desktop and the phone version, which share the same card. Hover over the question mark at XNatura → Overview → The numbers that matter.

The explanations next to the indicators on the Risks and Opportunities page had already been rewritten and checked over the past few days, but since these are texts describing environmental data used to take real decisions we wanted a third pass, this time purely for verification. Eight independent reviewers went back over all one hundred and sixteen explanations, statement by statement, without seeing the outcome of the previous pass: ninety-one were confirmed correct and twenty-four were corrected, along with five indicator labels. Among the most significant corrections: a susceptibility scale that runs from one to five rather than from zero, where zero does not mean negligible risk but missing data; the ranges for hot days and fire danger days, where the same boundary value must not be counted twice; and the snow figure, which is the maximum of the monthly values rather than a daily maximum. We also corrected two inaccuracies the previous pass had introduced, which is precisely why an independent re-check is worth doing. In one case we removed the soil depth quoted in an explanation, because it was not confirmed by the available sources and a precise but unverifiable number is worse than no number at all. In the same pass we removed the reference period given in the drought index explanation, because it contradicted the one stated in the Methodology panel on the same page: reading two different year ranges on one screen is more harmful than reading none, while the rest of the explanation is unchanged. All the corrections are already live in Italian, English, Spanish, French and German, so the explanation you read is the same whichever language you use the platform in. Hover over the question marks at XNatura → Monitoring → Risks and Opportunities.

In the social impact section of the Overview, and more generally wherever video content appears on the platform, the films were delivered through temporary addresses with an expiry date, inside an external player frame that had to load before the video itself. The mechanism worked, but it added a wait to every playback and required those addresses to be renewed periodically. We have moved the whole library to direct delivery of the video files: no player to load first and no links that expire. In practice you press play and the video starts, with a noticeably shorter initial wait, and scrubbing back and forth along the progress bar responds far more immediately. The starting point of films that begin at a specific moment has been kept identical, so you will not find videos restarting from the beginning where they used to jump to the right minute. Playback on phones and tablets benefits most of all, because that is where the initial wait was most noticeable. You need to do nothing and nothing changes in appearance: the same videos, in the same places, simply readier. A few particularly long films will be moved over the coming days and in the meantime keep working exactly as before. You will find them at XNatura → Overview, in the social impact section.