I need to learn what?

Harbaugh (zhongwen.com) works for me:

沙 = "Visible when little water", i.e. sand is what becomes visible when there is little (少, phonetic) water (氵).

漠 = "Without water" ( 莫 = phonetic, by itself a negating adverb that Harbaugh glosses as 'not', with the explanation "Sun 日 disappearing behind bushes 茻 (altered)" )

I remember being a bit confused about 沙, because I learned 砂 as "sand" and saw it way more frequently than 沙 for a while. But that was in a Japanese context, so... Still, I never did understand the difference. Maybe I'll decide to look into that at some point.

I think this is where learning compounds as opposed to just individual characters comes in: "shamo de sha" versus "shatang de sha" (latter with the stone rather than water radical, i.e. the 'sugar' sha, a word that will obviously be often see on sachets in restaurants, and in ingredients etc). Mnemonically, water (or the lack of it) is probably more associable with deserts than sugar. Anyway, I doubt that any practical dictionary could do much more than this to distinguish "synonyms" of 'sand', eh!

That seems to indicate maybe that 沙 is "sand" and 砂 is more like "granule".

Ah, interesting. Thanks for that, it does make things easier to remember.

For some reason the UI in zhongwen.com and I just do not get along....

I was going to check in a few C-E and E-C dictionaries for translations such as 'granule', but didn't have time earlier and then promptly forgot about it LOL. Looking now through a few C-E dictionaries or sections, and the E-E-C editions of the Longman Lexicon and Oxford Advanced Learner's Dictionary, it seems that the closest general meaning of 砂 (other than 'sand' LOL) would be 'grit' or 'gravel'.

Yes, the UI of zhongwen.com could perhaps be improved, but at least it's reasonably fast to load and click back and forth through. The genealogical trees are undeniably a work of genius, but in the book version I'd've opted for a more straightforward main means of look-up (that is, lost a few of the crossreferences or linking items and produced a more streamlined work) - the trees must've been hell to design and proofread (there are lots of little numbers and crossreferences in the book that aren't needed in the online version). So I guess the genius of that original paper design impacted both positively and negatively on the electronic design, both enabled it yet also limited it.

One thing I find slightly revealing (though maybe a bit of a shame) is that Harbaugh opted to not use his 182 字谱 elements, that (paraphrasing here!) "generalize and extend the radical system to all components whether semantic or phonetic", as the primary indexical way (like they are in the book) into the online version, but obviously those elements are still the backbone of, and found throughout, the online trees (and one can sort of still just about use them, if one is prepared to almost perversely click through a fair number of the arrows either side of the 字谱 hanzi at the top of each tree like they were pages in the book LOL); see the first paragraph of page 1, and the chart on page 7, of:

( http://yalepress.yale.edu/yupbooks/book.asp?isbn=9780966075007 > ) http://yalepress.yale.edu/yupbooks/languages/pdf/Harbaugh_intro.pdf .

Anyway, it's very generous of Harbaugh to have made it available for free online!

According to Wang Li et al. (2000). 《王力古漢語字典》. Beijing: 中華書局,

砂 is a later arisen character with two definitions:

1. A vulgar variant of 沙, which primarily means "sand" or "desert"
   
2. An abbreviation for vermilion paste